The proportion of Medicaid beneficiaries enrolled in managed care plans has increased dramatically over the past decade. Whereas in 1991 only 9.5 percent of Medicaid beneficiaries were enrolled in some form of managed care, the proportion had grown to 55.6 percent by June 1999, at which point there were 17.8 million Medicaid managed care enrollees (1). Recent headlines announcing the withdrawal of major managed care plans from state Medicaid markets suggest the potential for significant disruptions to the care received by millions of Medicaid beneficiaries (2).
Gathering reliable data on the effects of these exits poses a significant challenge. However, the extent to which problems will ensue is likely to be driven by several factors, including aspects of the delivery system in each state and the health care needs of the Medicaid population. This column considers the risks that exits by managed care plans pose for Medicaid enrollees with mental health and substance abuse conditions. Case studies are used to explore whether the impact of plans' exiting the market is driven in part by state approaches to the delivery and financing of services for this population.
During the rapid expansion of Medicaid managed care, important changes have occurred in both the nature and the number of plans serving Medicaid beneficiaries. The Balanced Budget Act of 1997 eliminated the requirement for plans to maintain a Medicaid enrollment of less than 75 percent—the "75/25 rule." In June 1997, 36 percent of Medicaid managed care enrollees were enrolled in plans that served only Medicaid beneficiaries (3).
Whereas 260 full-risk plans served the Medicaid population in 1995, the number had grown to 339 by 1997 (3). Although these figures depict a marked net upward trend in managed care participation, they obscure the fact that some plans enter state Medicaid markets while others exit every year. In 1996, only 16 commercial plans exited the market while 54 entered; in 1997, 36 commercial plans exited and 44 entered (3). However, in 1998 several large commercial plans, including PacifiCare, Blue Cross and Blue Shield Associations, Tufts Health Plan, Oxford, and Aetna/U.S. HealthCare, grabbed headline attention when they decided not to renew their Medicaid—or Medicare—contracts in several states (2).
The chief reasons cited by plan executives and others for plans' exiting the market are Medicaid rate reductions, inadequate risk adjustment, heightened competition in the Medicaid market as more plans have entered the market, changes in participation rules—for example, from voluntary to mandatory enrollment—that have burdened plans with risk-selection problems, new marketing restrictions, and the more stringent quality-assurance and monitoring requirements that are now often included in Medicaid contracts (4).
The data cited above suggest that exits by plans did not result in an overall decrease in access to managed care plans given the large number of plans that entered the market over the same period. However, looking at these numbers only in the aggregate may lead to overly optimistic conclusions. For example, in regions with fewer available plans, Medicaid enrollees may have trouble gaining access to treatment when a plan serving the area exits the market. In addition, the fact that the array of plans from which enrollees may choose is in a constant state of flux raises concerns about disruptions to care.
Although alternative plans are usually available, patients enrolled in exiting plans need to assess their options, select a new plan—or be assigned to one—and become acclimated to the various features of the new managed care system. In many cases, their relationships with providers may be disrupted and treatment may have to be adjusted because of changes in utilization-management techniques or general approaches to treatment. Thus a plan's exiting the market may be a particular concern for individuals with mental health or substance abuse conditions, who may be the least able to navigate a new health system or to adjust to a new approach to care management. In addition, continuity of care and stable provider relationships may be especially critical to the effectiveness of treatment for these patients (5).
Getting beneath the numbers of commercial managed care plan exits to fully understand the local implications for Medicaid enrollees with mental health or substance abuse problems may be especially difficult, given the variation across state Medicaid programs in the organization of these services (6). Generally speaking, states structure their services in one of three ways: integrated programs, in which behavioral and physical health services are both covered under a single capitation rate paid to the managed care plan; carve-out programs, in which behavioral health services are covered by a separate plan that is paid to manage only these services; and fee-for-service programs, in which behavioral health services are not covered by a managed care plan at all but rather are delivered on a fee-for-service basis.
We describe situations in three states in which managed care plans have exited the Medicaid market within the past three years. We selected the states to span the range of approaches to delivery of behavioral health services, in order to illustrate how the impact of plans' withdrawals on enrollees who are using these services may vary as a function of program type. We settled on this particular combination of states because the plans exited recently and the states enroll at least some share of their Medicaid populations in managed care on a mandatory basis. The information was gathered in telephone interviews with Medicaid officials in each state.
Missouri's integrated program
In 1998 and 1999, several managed care plans exited Missouri's Medicaid market. In the eastern region of the state, Humana withdrew at the end of its contract and Prudential was purchased by HealthCare USA. In the central region, Blue Choice exited at the end of its contract and Gencare left midcontract because it had been purchased by United HealthCare. The only plans serving the northwestern region—Blue Advantage Plus and Community Health Plan—withdrew from the Medicaid market as well, forcing their enrollees into fee-for-service arrangements.
State officials report that the process of having enrollees make the transition into new plans in the eastern and central regions went relatively smoothly. The officials attribute this success to several factors, including the number of managed care plans that entered the market at or around the same time; the large degree of overlap among the provider networks serving plans that were exiting, currently available plans, and those that were entering; and the open enrollment period of approximately two months that was given to enrollees to select a new plan.
Missouri allows enrollees receiving mental health or substance abuse services at the time a plan exits to continue with their current provider, even if continuation entails going out of the new plan's provider network. The providers in the networks of exiting plans also work to familiarize the care managers in the new plans with the treatment profiles of the mental health and substance abuse population. In the case of the enrollees in the northwestern region who had to return to a fee-for-service environment, no significant disruptions to behavioral health care were reported by state Medicaid officials, even though several providers had to be enrolled as fee-for-service providers under Medicaid.
Oregon's carve-out program
Oregon maintains a statewide carve-out program that contracts with multiple vendors that provide only behavioral health services. The state also has an integrated program that operates in selected regions. In those regions, enrollees who do not opt for the integrated program are automatically enrolled in the carve-out program. Over the past few years, Oregon has seen much activity in terms of Medicaid managed care plans entering and exiting specific service regions. In addition, in 1996 and 1997, three plans—PACC HMO, PacifiCare of Oregon, and QualMed —left the Oregon market altogether (3). Among other reasons for the exits, one state official reported that many dissatisfied providers in rural areas had left plans to start independent practice associations, forcing plans with depleted networks to leave the market.
Although state officials report that plans' exiting the market has generally not caused problems for enrollees in the integrated behavioral health program, the impact may vary somewhat across the state as a function of the degree of overlap among provider networks. There has been much movement among plans in the coastal and southern regions of the state, but few problems have resulted, because most physicians contract with multiple Medicaid managed care plans. On the other hand, in the more rural central and eastern regions, many plans prohibit their physicians from contracting with more than one plan. As a result, although there has been more plan stability in these areas, disruptions to the patient-physician relationship may be more likely if plans leave the market.
Vendors contracting with the state's carve-out program have been much less likely to leave the Medicaid market in Oregon. This stability may be attributable in part to the fact that these mental health and substance abuse plans serve broader regions of the state. In terms of the impact on carve-out enrollees of physical health plans leaving the market, it is worth noting that because some behavioral health services are provided in the primary care sector, disruptions to treatment relationships may result, even for carve-out enrollees, if there is insufficient overlap among primary care provider networks.
New Jersey's fee-for-service program
Three managed care plans—Garden State Health Plan, Oxford Health Plan, and HIP Health Plan of New Jersey—exited the New Jersey Medicaid market in 1997 and 1998. HIP Health Plan left the state entirely in September 1998 after becoming insolvent. The American Preferred Providers Plan has since left the Medicaid market as well.
Although consumers who were previously enrolled in these plans were randomly assigned to other plans with preexisting contracts, there was little disruption to mental health and substance abuse services because these services are not covered by managed care plans but are instead provided on a fee-for-service basis. Again, any behavioral health care received in the primary care sector may have been disrupted. Although the fee-for-service approach may raise questions about the coordination of care between physical and behavioral health services, these types of problems are not related to the issue of plan exit and thus are beyond the scope of this column.
These three case studies suggest that carve-outs and fee-for-service arrangements offer some protection against disruption to behavioral health specialty treatment for Medicaid beneficiaries in the event of managed care plans' exiting the Medicaid market. However, exit by a specialty mental health or substance abuse entity, through either termination or nonrenewal of a contract, is also a possibility. The exit of specialty entities, particularly through contract termination, has been less common than the exit of physical-health or integrated managed care programs in general, but it has occurred. For example, in Montana, Magellan Behavioral Health ended its contract after conflicts with the state over problems with claims payments and data collection, along with what the state claimed were poor relationships between the plan's providers and its management (7,8,9). Also, Massachusetts switched carve-out vendors when reprocuring its carve-out contract in 1996.
The rarity of termination of carve-out contracts is probably attributable to several factors. One possible explanation is that states use detailed requests for proposals and contracts in an attempt to clearly outline the responsibilities and expectations of both parties in advance. Another such factor is the presence of state regulations that govern the financial reserves necessary for plans to hold behavioral health carve-out contracts.
Although both contract termination and contract nonrenewal are likely to be disruptive, nonrenewal is potentially less disruptive given states' ability to plan ahead to ensure a smooth transition from one contractor to the next. For example, states may expand their networks to include providers who are currently treating enrollees, give enrollees a transition period during which to switch to a network provider, or extend the open enrollment period to give enrollees more time to select a new plan. It is also worth noting that the level of disruption created by a plan's exit may be influenced by the extent of involvement of public-sector providers. For example, a new carve-out vendor's decision to restrict the volume of referrals to public providers could both threaten the continued viability of the providers and harm established patient-provider relationships.
In many ways, the issues raised by the exit of managed care plans simply highlight the potential problems associated with managed care competition itself. In other words, it is not clear whether the disruptions that might occur in the wake of a plan's loss of a private contract in a competitive bidding cycle are different from those that might arise after a plan's withdrawal from the Medicaid market. However, as Medicaid enrollees have been shifted into managed care, clinicians and policy makers alike have expressed particular concerns about the implications for patients with mental health or substance abuse problems.
Although the examples presented in this column suggest that there are ways of easing the transitions that are necessary in the wake of plan exits, it is critical to note that the types of disruptions that may occur for Medicaid enrollees who need behavioral health services may not necessarily be reported to or documented by the state. In many ways, the difficulty of monitoring the care received by exiting plans' enrollees is itself one of the risks of exits by plans.
As we have argued, the impact of exits on Medicaid beneficiaries seeking mental health or substance abuse services is likely to be driven in part by state Medicaid programs' organizational approaches to coverage and delivery of these services—a dimension in which there is currently great variation (6). Additional important factors include the degree of overlap among local plans' provider networks, the number of enrollees receiving behavioral health services in the primary care sector, transition policies affecting enrollees, and the extent to which new plans have entered the area during the same period.
This work was supported by the Center for Health Care Strategies, Inc. The center was made possible by a grant from the Robert Wood Johnson Foundation's Medicaid Managed Care Program. Funding was also provided by the National Institute for Drug Abuse through grant P50DA-10233 from the Brandeis-Harvard Center on Managed Care. The authors thank Rob Cenczyk, Julie Donohue, Domenic Hodgkin, Ph.D., Frank McCorry, Ph.D., and Aileen Rothbard, Sc.D., for their helpful comments and assistance.
Dr. Huskamp is affiliated with the department of health care policy at Harvard Medical School, 180 Longwood Avenue, Boston, Massachusetts 02115 (e-mail, firstname.lastname@example.org). Dr. Garnick and Dr. Horgan are affiliated with Brandeis University in Waltham, Massachusetts. Ms. Hanson is with the Henry J. Kaiser Family Foundation in Washington, D.C. Howard H. Goldman, M.D., Ph.D., and Colette Croze, M.S.W., are editors of this column.