Number A-11 in Series, "Issues and Options for States"
The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.
The number of Americans enrolled in managed care has grown dramatically during the 1990s, as private and public purchasers of health care turn to managed care as a way of providing more cost-effective delivery of health services. Managed care programs work toward these goals by changing how providers are paid, developing select provider networks, establishing protocols for appropriate care, and offering enrollees financial incentives to receive care from specific providers. The private sector has already achieved substantial savings through managed care. State Medicaid programs are increasing their use of managed care in the hope of achieving similar success.
This brief reviews the questions state Medicaid programs must answer for themselves in order to maximize their chances of increasing savings while preserving or increasing access to care. To provide a context for this discussion, we begin by briefly reviewing recent Medicaid enrollment trends, state objectives in turning to managed care for their Medicaid programs, and federal rules constraining how states structure those programs.
Medicaid Managed Care Enrollment
Over the period since 1991, Medicaid enrollment in managed care has grown much more rapidly than total Medicaid enrollment (
figure 1). Overall, Medicaid enrollment increased from 28.3 million to 33.2 million between 1991 and 1996, while managed care grew from 9.5 percent to 40.1 percent of total Medicaid enrollment during the period.
1
Forty-nine states now rely on some form of Medicaid managed care plan. But not all these states rely on capitation—paying the plan a fixed sum per enrollee for a required set of services—which transfers full financial risk to the plan. When state Medicaid programs first began using managed care in the early 1980s, the predominant form of managed care was Primary Care Case Management (PCCM). PCCMs match beneficiaries with gatekeeper providers, who take primary responsibility for coordinating beneficiaries' care and receive a monthly fee of about two to three dollars in exchange. Other services are paid for on the traditional fee-for-service basis. PCCMs emphasize the access-enhancing role that managed care can sometimes play, but have resulted in only modest savings.
In the early 1990s some states started to become more active in seeking Medicaid savings by moving toward capitation plans, of which there are two types. Full-risk Health Maintenance Organizations (HMOs) are fully capitated for a wide ("comprehensive") range of required benefits across preventive, primary, and acute care services. Prepaid Health Plans (PHPs), which are mostly clinics or large group practices, are typically only at risk for ambulatory services (which are frequently restricted to mental health services and dental care), but not for a comprehensive set of benefits.
The widely shared perception is that states have been moving in the direction of full capitation for several years. In fact, the growth of fully capitated Medicaid HMOs did not explode until 1995–1996, when full-risk HMO enrollment grew by over 60 percent, more than five times the rate of increase of PCCMs and PHPs (figure 2).2 Even so, about one-third of Medicaid managed care enrollees in 1996 still remained in PCCM arrangements, and 13 states had PCCM programs as their only type of Medicaid managed care.3 Those states that continue to rely on or expand their PCCM programs do so for three reasons. Either they do not have viable managed care plans willing to accept full financial risk in most areas of the state. Or providers have successfully used political pressure to block the adoption of capitated payment systems within Medicaid. Or, states may view PCCM programs as preferable to HMOs for certain populations, such as people with disabilities.
States' Objectives and Federal Rules
States have two objectives in moving toward managed care for their Medicaid enrollees. The first reason is to reap savings, by changing the incentives that insurers and providers have faced historically to overprovide services. Second, managed care within Medicaid also offers the potential of enhanced beneficiary access to and continuity of care.
Historically, Medicaid beneficiaries' access to care has been threatened in the fee-for-service environment, at least partly because states have set fees well below those of other insurers, resulting in beneficiaries tending to rely on the emergency room and a subset of clinics that treat predominantly Medicaid patients.
Many states have been reluctant to increase fee-for-service payment rates in order to improve access because of fears that increasing payment rates will not increase the number of Medicaid providers but simply increase payments to those already serving Medicaid clients.4 States are turning, rather, to requiring managed care plans to develop mechanisms for attracting and retaining an adequate number of providers. States prefer this to simply increasing provider payment rates because a state's expenditure risk is limited and plans' performance can be monitored. In addition, by identifying plan providers who have agreed in advance to accept specific Medicaid enrollees as patients, the expansion of managed care could improve beneficiaries' access to care.
Despite the potential advantages to Medicaid managed care, until recently federal law had not allowed states to unilaterally mandate that beneficiaries enroll in a managed care plan and give up their ability to choose a provider. To make managed care more than a voluntary option for beneficiaries, states had to seek waivers. Two types of waivers have been granted: Section 1915(b) "Freedom of Choice" waivers and Section 1115 research and demonstration waivers.
A Freedom of Choice waiver allows states to mandate managed care for Medicaid enrollees, but beneficiaries still need to have a choice among health plans. As of January 1997, 96 Section 1915(b) waivers had been granted in 42 states, with some states holding waivers for more than one geographic area within the state.
Section 1115 waivers allow states even more flexibility. Because they are viewed as demonstration projects, the federal government can approve states' requests to test new approaches to benefits, services, eligibility, delivery systems, and program payments. These waivers are often designed to save money by making managed care mandatory for current enrollees and using the resulting savings to expand Medicaid eligibility. In addition, these waivers allow states to limit the types of plans available to beneficiaries to a greater degree than Freedom of Choice waivers. Sixteen states have received approval to run comprehensive Section 1115 demonstration projects. Of these, 10 have been implemented—in Arizona, Oregon, Tennessee, Hawaii, Rhode Island, Minnesota, Delaware, Ohio, Oklahoma, and Vermont.
The need to seek these waivers in order to implement mandatory managed care programs has been altered by the 1997 federal budget legislation. That legislation allows states to make Medicaid managed care mandatory without waivers under certain conditions.
Key Questions Facing States
Within the overall trend of state Medicaid programs toward managed care, there is wide state-by-state variation.
Table 1 shows the distribution of the 11.7 million Medicaid managed care enrollees by state and the relative importance of full-risk HMOs and PCCMs in each state as of June 1996.
5 Despite the dramatic growth in managed care enrollment when viewed from a national perspective, many states still have very few Medicaid beneficiaries in managed care.
The three states with the largest share of their enrollees in Medicaid managed care programs are among those that have implemented major Section 1115 waiver programs (Arizona, Tennessee, and Hawaii). Their high penetration rates are in stark contrast to the single-digit managed care penetration rates at the other end of the range (in Louisiana, Maine, Mississippi, Texas, and Wyoming).6
Based on our discussions with federal and state officials, health care providers, and representatives of beneficiaries, we have identified a series of critical questions that states need to address as they move their Medicaid programs in the direction of mandatory full-risk capitation managed care.
Which types of beneficiaries should be required to enroll in managed care? This question is important to states because it influences the extent of savings that can be expected, as well as the difficulties managed care plans are likely to have in serving the Medicaid enrollee population.
In most states, Medicaid managed care enrollment has been limited to nondisabled adults and children. This includes AFDC recipients (i.e., adults and children meeting July 1996 eligibility criteria), pregnant women, and low-income children. Managed care plans in the private sector typically treat relatively young and healthy individuals and generally have limited experience treating chronically ill patients. Given this experience, when states turn to managed care for their Medicaid programs they typically exempt the elderly, persons with disabilities, and those eligible through medically needy options (i.e., persons whose health care costs are extraordinarily high in comparison to their incomes).7
Nondisabled adults and children are by far the largest group of Medicaid beneficiaries, accounting for 73 percent of all Medicaid beneficiaries in 1994—quite large enough to give managed care plans a strong incentive to establish provider networks, set up administrative systems, and develop marketing strategies. However, by limiting enrollment to these beneficiaries, states also have limited themselves to the lowest cost group—accounting for only 32 percent of Medicaid expenditures in 1994.8
Elderly and disabled beneficiaries—the remaining 27 percent of Medicaid beneficiaries—have much higher costs and, as a result, offer the greatest potential savings from care management. But states often find it difficult to identify and encourage the development of managed care plans that can adequately treat chronically ill and disabled populations—individuals with, for example, severe developmental disabilities, severe mental illness, and AIDS. Initially, Medicaid could even have to cover some of the costs associated with developing these new and complex types of plans. While several states have debated the feasibility of enrolling these groups in managed care, few are actually proceeding down this road.
Medicare-Medicaid dual eligibles are difficult to include in managed care for somewhat different reasons. First, some states would like to create managed care that extends across Medicare and Medicaid in order to increase service continuity and reduce cost-shifting.9 However, there are enormous complexities involved in coordinating across the two programs, aggravated by the fact that not all dual eligibles receive the same benefits from Medicaid.10 Second, this is a population that often requires long-term care. The complex interface between acute and long-term care systems has yet to be satisfactorily navigated outside small demonstrations.
Third, Medicare pays for most of the acute care services, leaving Medicaid little financial leverage over physicians, who control health care referrals and costs to a significant degree. Moreover, even if beneficiaries are enrolled in the same plan for Medicare and Medicaid benefits, they are free to disenroll from their Medicare plan at the end of any month, according to federal law. States may not override this federal provision for dual enrollees.
What types of services should Medicaid managed care plans be required to provide? Medicaid programs have historically paid for a broader range of services than many private managed care plans. For example, some Medicaid programs provide extensive coverage for substance abuse treatment, mental health care, rehabilitation, home and personal care, nursing home care, and case management. If managed care plans take on responsibility for these services and are either unable to deliver them or view them as "medically unnecessary," the move to managed care could limit beneficiary access.
One response to this access problem has been to "carve out" mental health and/or substance abuse services and contract with specialized managed care plans to provide these services or leave them in Medicaid fee-for-service. But carve-outs are not without their problems. If a set of services is not explicitly included under a health plan's capitated obligation, the plan has an incentive to interpret its coverage narrowly and shift costs to the fee-for-service realm or other managed care plans. In such cases, states run the danger of overpaying plans. Even without these incentives to shift costs, states may find it difficult to clearly define a managed care plan's contractual obligations. Moreover, it may be difficult to track the specific services that are provided by managed care plans. Under fee-for-service Medicaid, states could track services from claims providers submitted for payment. However, because managed care does not link payments to claims, states must require plans to report detailed service information for individual patients. Plans may have difficulty providing this information in a form useable by the state.
Do states need to ensure the continued existence of traditional Medicaid providers as a part of managed care plans? Here states have three choices. They may enroll Medicaid beneficiaries in capitated plans that include mostly private subscribers, termed "mainstream plans," they may rely primarily on plans that are formed predominantly to seek Medicaid and Medicare business, or they may use a combination.
If the goal is to move Medicaid beneficiaries into mainstream health plans and give them access to a new set of providers, then the state is less likely to require that public hospitals, neighborhood clinics, and local health departments be included in managed care plans. However, if policymakers believe that these providers are best suited to continue treating Medicaid patients, states could require managed care plans to include such providers in their networks as a condition of receiving a Medicaid contract. This decision may be complicated by the reality that these providers are the major sources of uncompensated care for the uninsured and need Medicaid patients and the associated revenues to survive financially. Some states may even encourage these traditional safety net providers to establish their own managed care plans and develop their own networks of providers.
How will the enrollment process work? Enrollment in managed care may be voluntary or mandatory. In a voluntary system, beneficiaries choose either to join an HMO or to stay in the fee-for-service system. HMOs may be attractive to beneficiaries because they offer a medical "home" and improved access to some services. But persons with prior provider relationships may avoid HMOs because they do not want to risk changing providers. In any case, it is important to ensure that beneficiaries understand their choice of delivery system and their responsibilities. If beneficiaries are enrolled in private plans and do not understand how they operate, public providers could be responsible for service costs when sick patients "incorrectly" show up at emergency rooms or non-plan providers for care.
Mandatory enrollment in managed care increases the obligation of the state to provide adequate access to services. Under most waivers, the federal government requires that the state offer a choice of managed care plans. Mandatory managed care programs typically give beneficiaries a period of time to select from a list of managed care plans. High rates of plan selection during a specified period are considered a sign that beneficiaries have understood their obligation to choose a plan and will know where to seek care when it becomes necessary.
Some states have already experienced difficulties in getting beneficiaries enrolled in managed care plans. One problem has been that states have allowed managed care plans to contact beneficiaries directly, which has led individuals to make choices without having all the information the state might reasonably want them to have. Some states have tried to correct this by hiring enrollment brokers to make the initial educational contacts with the Medicaid population. However, this has not always increased the proportion of enrollees who choose a plan during the selection period.
When individuals do not choose a plan within the allotted time, states in which enrollment is mandatory must automatically assign them to a plan. This process—termed "auto-assignment"—has the major drawback of eliminating the influence of enrollees' preferences on competition among plans, however. Without such competition, there is less market pressure to encourage plans to maintain service quality at acceptable levels. States can counteract this disadvantage by establishing a process that channels enrollees to plans that meet specific cost, quality of care, or service standards.11 By rewarding plans in this way, even substantial auto-assignment can still produce quality competition among plans.
How will managed care plans be paid? Once the state decides which beneficiaries and services will be included in a capitated managed care plan, the monthly payment must be set so that the state realizes some savings relative to what would have been paid under fee-for-service Medicaid. States are very aware of their need to build capacity and do not seem to be making major attempts to slash rates. Typically, states expect savings in the range of 5 to 10 percent. They use this assumption as a basis for rate setting, negotiations, and target setting in bidding arrangements with managed care plans.
As enrollment in managed care increases in a state, the ability to base capitated rates on that state's experience with fee-for-service Medicaid will shrink because the Medicaid program's fee-for-service claims experience will begin to disappear. This may lead to rates based on direct negotiations with plans or competitive bidding. Since Medicaid is not the dominant payer in many markets, the need to negotiate or solicit bids has led some states to align with other health care purchasers.
No matter how rates are set, states need to be concerned that managed care plans have limited incentives to enroll the patients with higher-than-average expected costs. For example, a state would not want to set the same rate for pregnant women as for children. If it did, health plans would have an incentive to skew their services so as to attract the lower cost pediatric patients. Capitated rate setting can become even more complicated if Medicaid requires managed care plans to enroll disabled or elderly beneficiaries. Adjusting rates for risk is crucially important in this connection. Unfortunately, methods for doing this are still quite rudimentary.
Setting rates for health plans is also complicated by the fact that certain providers are eligible to receive special payments from Medicaid. These special payments typically reflect disproportionate-share payments for hospitals serving more than their share of low-income clients or cost-based reimbursement for some community health centers. In such cases, states must decide if these payments are to be incorporated into the capitated rate paid to the health plans, or if providers will continue to receive them directly from the state.
How will states monitor plan performance? States will want to know they are getting what they paid for, which requires some mechanism to oversee plan performance. To monitor service utilization and quality of care, some states may require plans to provide detailed information on client encounters. Other states may choose to rely on summary measures of plan performance—such as immunization rates or the frequency of hospitalizations for conditions that would not require such care if properly monitored on an outpatient basis. Many states require plans to document internal quality control procedures and general plan management. States also look for some measure of the adequacy and appropriateness of the provider network. However, measuring these factors is difficult because providers have patients outside Medicaid and have contracts with many different managed care plans, including non-Medicaid plans.
Even if no quality or access problems are detected, many states will scrutinize plan profitability on a continuing basis. High profit rates may indicate that capitation rates are higher than necessary to provide adequate care. High profit rates may also imply that some of the spending that previously went for health care services has been shifted to finance administrative costs or bonuses to plan managers and providers.
Conclusion
Despite the broad interest in managed care among Medicaid programs, the amount of Medicaid spending channeled through managed care is still limited. A variety of factors are responsible. First, most states have excluded their most expensive beneficiaries from these plans and are still paying for long-term care services on a fee-for-service basis. Second, many states continue to rely on PCCM plans that are not based on capitating provider payments, leaving only about 7 percent of Medicaid spending (as of 1994) managed by fully capitated health plans.
12
As in the private sector, the movement to managed care within Medicaid represents a dramatic change from business as usual. And the demands placed on managed care plans may be far greater within Medicaid than in the private sector. Ultimately, Medicaid managed care plans may need to treat more chronically ill and severely disabled individuals and provide a far broader range of services than has been required by private buyers. At the same time, managed care plans that choose to participate in Medicaid will have to operate with relatively low capitation payments (reflecting Medicaid's historically low fee-for-service rates) and with intense public oversight.
Given these requirements, will Medicaid be able to attract sufficient and appropriate managed care plan capacity? Will enough mainstream plans with the ability to treat all types of Medicaid enrollees agree to participate at rates that allow the program to achieve savings? Or will the program become dependent on plans that contain predominantly Medicaid enrollees and offer access to the same safety net providers who have historically treated this population?
If the program's low rates prevent all but a few plans from competing for Medicaid patients, the plans that do participate may be able to use their market power to force rates up, thus limiting program savings. Although recent growth in Medicaid managed care enrollment has been rapid, it remains to be seen if the Medicaid managed care market will develop in a way that allows most beneficiaries to enroll in these plans while preserving access to quality care and lowering program costs.
Notes
1. These data include managed care enrollees in all 50 states, the District of Columbia, and Puerto Rico. There were 802,000 Medicaid managed care enrollees in Puerto Rico as of June 30, 1996. Excluding Puerto Rico brings the number of Medicaid beneficiaries enrolled in managed care down to 12.5 million.
2. The data in figure 2 show that the sum of enrollment in full-risk HMOs, PHPs, and PCCM programs exceeds the total managed care enrollment reported in figure 1. This is because of double-counting in figure 2, i.e., some people are enrolled in more than one type of managed care.
3. These 13 states are Alabama, Arkansas, Idaho, Kentucky, Louisiana, Maine, Mississippi, New Mexico, North Dakota, South Carolina, South Dakota, West Virginia, and Wyoming.
4. Perloff, J.D., Kletke, P., and Fossett, J.W., "Which Physicians Limit Their Medicaid Participation, and Why?" Health Services Research, April 1995, vol. 30, no. 1, pp. 7–26.
5. Of the 13.3 million Medicaid managed care enrollees shown in figure 1 for 1996, 0.5 million are in plans that only provide a limited set of services (generally, dental or mental health care), 0.3 million are in PHPs, and 0.8 million are residents of Puerto Rico. All three of these categories are excluded from table 1.
6. Alaska, Nevada, South Carolina, and Vermont reported no enrollment in full-risk HMOs or PCCMs in 1996. Both Nevada and South Carolina did have enrollees in PHPs.
7. The major exceptions to this are Tennessee and Arizona, which enroll virtually all of their Medicaid beneficiaries in mandatory managed care.
8. These statistics exclude disproportionate share payments. David Liska (1997), "Medicaid: Overview of a Complex Program," New Federalism: Issues and Options for States, series A, no. A-8, Urban Institute, May 1997.
9. "New England States Initiative for Dually Eligible Persons: Concept Paper," mimeo, April 1996. This concept paper outlines a research and demonstration proposal to the Health Care Financing Administration in which six states, Connecticut, Massachusetts, Maine, New Hampshire, Vermont, and Rhode Island, proposed a joint waiver in order to integrate care and combine Medicare and Medicaid funding. Since then, the states have chosen to pursue federal waivers separately.
10. For some dual eligibles (about 13 percent of elderly dual eligibles), Medicaid pays only the Part B premium or the Part B premium plus deductibles and copayments. Others are eligible for the full range of Medicaid services, including prescription drugs and long-term care services. An additional complexity is that one-quarter of dual eligibles reside in institutions (e.g., nursing homes or group homes).
11. Given how new this process is, there is still disagreement over the relative costliness of auto-assignees to managed care plans. On the one hand, they may not choose a plan because they believe their needs for services are low, suggesting that plans would benefit financially by having them as enrollees. On the other hand, their unwillingness to choose a plan may be a signal that they are likely to be less-compliant patients and, therefore, difficult and expensive to treat within the constraints of managed care.
12. Lewin-VHI (1995), States as Payers: Managed Care for Medicaid Populations (Washington, D.C.: National Institute for Healthcare Management), February. (Based on the growth in managed care enrollment and the fact that it is occurring predominantly among full-risk HMOs, in 1996 the share of Medicaid spending based on capitated payments could be almost twice as high as in 1994.)
Tables and Figures
About the Authors
Stephen Zuckerman is a principal research associate at the Urban Institute's Health Policy Center. His primary research interests include health system reform, physician payment, and hospital costs and efficiency.
Alison Evans is a former research associate in the Health Policy Center. She participated in several case studies within the Assessing the New Federalism project and analyses of Medicare budget issues.
John Holahan is director of the Health Policy Center. His primary interests include health system reform, health care cost containment, Medicare, and Medicaid.
The series is dedicated to the memory of Steven D. Gold, who was co-director of Assessing the New Federalism until his death in August 1996.
PDF