Status of the Medicaid competition demonstrations

Abstract

In 1982, the Health Care Financing Administration approved funding for demonstration programs in six States to test a variety of alternative delivery strategies for Medicaid recipients. A number of innovative health service delivery features have been used in these programs, including competition, capitation, case management, and limitations on provider choice. These strategies have been tried in order to address the key Medicaid problems of cost containment and access to appropriate and high quality care. This article provides an overview of how the demonstration sites have approached the task of designing, developing, and implementing their various programs.

Introduction

In 1982, the Health Care Financing Administration (HCFA) approved funding for demonstration programs in six States. The programs were to test a variety of alternative delivery strategies for Medicaid recipients. To address the key Medicaid problems of cost containment and access to appropriate and high quality care, the programs have used a number of innovative health service delivery features including competition, capitation, case management, and limitations on provider choice. The programs have incorporated these features into several different types of organizational arrangements in order to test a number of assumptions about how the delivery system can be effectively changed.

In the fall of 1983, HCFA awarded a contract to a consortium of researchers under prime contract with the Research Triangle Institute¹. The researchers were to conduct a 4-year evaluation of these demonstration programs. This evaluation is designed to perform a comprehensive assessment of the demonstration strategies including implementation and operational issues as well as program outcomes. The evaluation plan includes both quantitative and qualitative components to accomplish this goal.

The analysis of program effects, based on such outcomes as cost, use, access, quality, satisfaction and provider participation will be conducted with primary and secondary data collected during operation of the programs. The evaluation team is also examining design, development, and implementation issues. This is primarily being done through a series of detailed, multiyear case studies carried out at each of the demonstration sites by personnel from Lewin and Associates, the American Enterprise Institute, and New Directions for Policy. This article describes the set of case studies performed in the sites during 1985-86. The final report of the evaluation is due in 1987.

Purpose of the demonstrations

The demonstration programs are exploring whether alternative approaches to providing care can respond to the many problems that have plagued the Medicaid program during its 20 years of existence. These problems include, but are not limited to, the following:

• Excessive rates of cost increases.

• Unnecessarily high rates of use for selected services.

• Inappropriate patterns of use such as reliance on the emergency room for nonemergency care; high rates of self-referrals to specialists; and “doctor shopping,” i.e., capricious changes in medical providers.

• Lack of access to providers offering continuity of care.

• Concern that available providers may not provide high quality care.

• Declining physician participation for such reasons as unreasonably low fees, delays in receiving payment on a timely basis, and administrative burdens in negotiating the payment system.

Many of these problems are interrelated and self-reinforcing, suggesting major structural reform must be explored in the Medicaid program. These demonstrations with critical elements of competition, capitation, and case management are among several delivery system reforms currently being evaluated by HCFA.

Competition has been included in these programs in order to attempt to bring providers into Medicaid who have traditionally had little or no involvement with the program. By expanding provider participation, problems of access can be addressed and, ultimately, costs may be contained and reduced by increased competition among new and existing providers. In response to the entry of new providers, traditional Medicaid providers are expected to modify their approaches to serving the Medicaid population in order to avoid loss of patients.

Financial risk-sharing with providers, in the form of prepaid capitated rates, is also being explored extensively in the demonstrations. The setting and payment of rates in advance to cover specified services gives participating providers concrete performance targets that they must meet to remain viable. By establishing rates below existing equivalent fee-for-service payment amounts, the demonstration programs can be assured of cost savings. Correspondingly, prepayment gives providers increased revenue predictability and improved cash flows. More significantly, providers come to recognize that substantial financial savings might be achieved by judiciously managing enrollee utilization, including limiting unnecessary use and substituting less costly services.

Case management attempts to address cost, use, and access problems by taking advantage of the pivotal role of the primary care provider as the point of access to the health care system. By linking and locking-in an eligible person to a primary care case manager, who can both provide and manage a substantial portion of a recipient's medical care, patterns of service use may be changed and access to appropriate care assured. The relationship of the provider to the Medicaid program and to the recipient can be structured in a number of ways, using contractual arrangements and risk-sharing approaches, which are designed to foster effective case management and to achieve program goals. The demonstrations represent a broad spectrum of planned variations intended to do this.

Background

An understanding of the development of the programs requires some background on the individual demonstrations. In Table 1, a synopsis of selected program characteristics is presented. The following is a brief description of each program.

Table 1. Selected demonstration characteristics, by demonstration site.

Demonstration site	Date of implementation	Type of enrollment	Organizational structure	Eligible population	Participating providers	Provider payment
California
Monterey County1	June 1983	Mandatory enrollment; choice of provider	Risk-assuming intermediaries which contract with primary care organizations and individuals	Categorically eligible and medically needy	Case managers are primary care providers, including physicians, clinics, and hospitals	Intermediary capitated Monterey—fee-for-service plus fee-Santa Barbara-capitation
Santa Barbara County	September 1983
Florida2	Planned	Voluntary enrollment	State contracts with prepaid plan	Supplemental Security Income—frail elderly	Hospital	Capitation
Minnesota
Dakota County	December 1985	Mandatory enrollment;3 choice of provider	State contracts with prepaid health plans or county (Itasca)	Aid to Families with Dependent Children, Aged, Blind, Disabled	Primary care organizations	Capitation for plans in Hennepin and Dakota and for county in Itasca
Hennepin County	December 1985
Itasca County	August 1985
Missouri
Jackson County	November 1983	Mandatory enrollment; choice of provider	State contracts with prepaid health plans and individual physicians	Aid to Families with Dependent Children	Plans include hospitals, individual practice association, neighborhood health centers, and individual physicians	Capitation for prepaid health plans; fee-for-service with fee for physicians
New Jersey	June 1983	Voluntary enrollment	State contracts with primary care organizations and individual physicians	Categorically eligible	Case manager must be primary care provider including health centers and physicians	Capitation
New York
Monroe County	June 1985	Mandatory enrollment; choice of provider	Intermediary which contracts with prepaid health plans	Aid to Families with Dependent Children, home relief, medically needy	Prepaid health plans	Capitation

¹Terminated March 1985.

²Three of four proposed modules terminated August 1984.

³Random assignment employed in Hennepin County.

Monterey

Operated by the Monterey County (California) Health Initiative, until its termination because of insolvency in March 198S, this demonstration provided a mandatory primary care case manager program. The primary care providers were paid on a fee-for-service basis with a case management fee; providers were not at financial risk for specialty and hospital care. Participating providers included physicians, health centers, and hospital outpatient departments. Enrollment reached 26,000 in December 1984, with 160 participating case managers.

Santa Barbara

This demonstration, operated by the Santa Barbara County (California) Health Initiative, under a prepayment contract with the State Medi-Cal agency, is a mandatory primary care case management program. The initiative contracts on a prepayment basis with such primary care providers as individual physicians, physician groups, and health centers; these groups are then responsible for providing primary care services and for authorizing specialty and hospital care. The program is fully operational, with approximately 21,000 enrollees and 125 participating case managers.

Florida

The State Medicaid agency originally planned four separate modules to develop alternative delivery systems using elements of prepayment, competition, and case management. Three of the four modules are no longer part of the demonstrations; they have either been terminated or undertaken by the State as nondemonstration programs. Planning for the fourth module continues and involves the development of a prepaid case-management program for the frail elderly. This program is expected to be implemented in 1986 in the Miami area; its objective will be to avoid nursing home placement by the provision and coordination of medical and social services.

Minnesota

The State Medicaid agency is conducting demonstrations in three counties: Dakota, Itasca, and Hennepin (Minneapolis). In Hennepin and Dakota (a suburban Minneapolis county), seven health organizations have entered into prepayment contracts to enroll eligible individuals who may select from any of the plans. In Itasca, a small rural county, the county receives a prepayment for each enrollee, and providers are paid on a fee-for-service basis with surpluses and deficits shared by the county and the providers. Enrollment is mandatory in the counties with the exception of Hennepin, where only 35 percent of the population will be randomly assigned to enrollment, and the remainder will stay in the traditional Medicaid program. Total enrollment as of July 1986 was approximately 11,700.

Missouri

The State Medicaid agency operates a mandatory enrollment program for Aid to Families with Dependent Children (AFDC) recipients of Jackson County including Kansas City. Most of the eligible population is enrolled with five prepaid health service organizations: two hospitals, two neighborhood health centers, and an individual practice association (IPA); these organizations are responsible for providing, or authorizing virtually all medical care. Approximately 20 percent of the eligible group are enrolled in the physician sponsor program (PSP) in which 55 primary care physicians are paid on a fee-for-service basis and receive a case management fee to manage care, including authorizing referral and inpatient services. Total enrollment is approximately 23,000.

New Jersey

This demonstration provides for the voluntary enrollment of Medicaid eligible individuals with primary care case managers, physicians and health centers, which are paid on a prepayment basis for each enrollee. The prepayment is structured to compensate the case manager for primary care services directly provided and to place the case manager at some financial risk for referral services. Operated by the State Medicaid agency, the program has been implemented in several counties and statewide implementation is planned. Enrollment in early 1986 was approximately 9,500.

New York

This mandatory program in Monroe County including Rochester, is managed by a county agency under a prepayment contract to the State Medicaid agency. This county agency, MediCap, contracts with a network model health maintenance organization (HMO) to provide case-managed services to the enrolled population. The provider members of the network include physician groups, neighborhood health centers, and hospitals; these members receive prepaid amounts to cover a broad range of medical services, which are either provided or arranged. MediCap is attempting to recruit other prepayment plans; enrollment had reached approximately 25,000 in February 1986.

Key terms

The variations in program design permitted and encouraged in the demonstrations make it important to clarify several terms used to describe certain program aspects across the demonstrations. Among the key terms and their definitions are the following:

Risk assuming intermediary

In two States, California and New York, intermediary organizations have contracted with the State Medicaid agency to manage the program in return for a fixed prepaid amount received for each eligible person enrolled in the program. These intermediaries, which provide no medical services themselves, are responsible for arranging service provision with area medical providers.

Prepaid health plans or organizations

These provider organizations enter into agreements to provide services directly with the State Medicaid agency or the risk-assuming intermediary organization. These organizations may range from conventional prepaid organizations, like HMO's, to other providers, such as hospitals and health centers; typically, these organizations are paid on a prepayment basis for a specified range of services.

Primary care case managers

Several of the programs have primary care case manager (PCCM's). In these programs, primary care physicians are formally designated as the case manager, i.e., gatekeeper, for a group of enrollees. PCCM's may have contractual relationships with the Medicaid agency, intermediaries, or prepaid health plans. PCCM's may be compensated by a prepaid payment for specified services or on a fee-for-service basis. In fee-for-service situations, the PCCM is usually paid a supplemental fee to perform case management duties. Some of the participating prepaid plans have elected to use the case management approach, and others have not.

Capitation

Programs have established prospective rates of prepayment, based on the average historical cost, to provide a specified set of services to eligible individuals. These rates, called capitation payments, represent the principal means of structuring risk sharing among the various organizations participating in each demonstration. The capitation rate may be set to include all Medicaid services, or the rates may be limited to a subset of services such as primary care services.

Status of selected issues

A number of critical demonstration program issues may be examined across the sites. These issues include: implementation, rate setting, enrollment, management information systems, provider participation, quality assurance, provider payment and risk-sharing, and administration and management.

Implementation

As shown graphically in Figure 1, the periods of time devoted to implementation have varied among the programs, but programs have consistently taken longer to implement than expected. These delays are particularly troublesome in time-limited demonstration programs. Program administrators report the tensions between getting started prematurely, on the one hand, and jeopardizing program credibility (support) and viability (funds) by being too deliberate, on the other hand. Attempting to satisfy conflicting interests of the Federal funding agency and the provider community puts severe countervailing pressures on those responsible for the programs.

Figure 1. Site development and implementation schedules: 1982-87.

The implementation period has been marked by enormously time-consuming efforts at consensus building and tradeoff negotiations with providers. Most important, these negotiations can result in program design changes that can fundamentally affect or alter the programs goals. For example, critics of Monterey have suggested that in the face of provider opposition, Monterey negotiated fee-for-service payment rates that were higher than conventional Medi-Cal rates; later efforts to tighten controls were strongly resisted. New Jersey granted a 1-year waiver of risk to early enrolling providers to break an impasse and begin operations. In New York, inability of provider groups to form risk-sharing entities limited competitive efforts to existing area HMO's. Missouri expanded its program, which initially was to encompass only prepaid plans, to include an entire set of Medicaid physicians who participate in the PSP. This expansion was necessary to defuse the providers' opposition without relinquishing program aims. As discussed earlier, the trade-off between getting started with existing Medicaid providers or attracting new providers has been another implementation dilemma faced by the demonstrations in a number of sites.

Interorganizational and intergovernmental relationships have also played a not-unexpected role in these public programs. Federal, State, and local officials have had varying expectations and commitments both to the overall program and selected program features. These concerns have surfaced in the design and the implementation phases. In some cases, disputes have arisen in purely technical areas like rate setting; in others, the concerns have followed more traditional jursidictional disputes, including where the locus of authority should reside to make changes necessary to implement the program. In several cases, municipalities operating health service facilities have been reluctant to embrace the programs because of fear of incurring undue risk. This fear is a consequence of municipalities' perceived uncompetitive positions and the vulnerable position in which their indigent care responsibilities may place them.

Enrollment

The enrollment process actually includes a number of related procedures:

• Consumer information and education.

• Provider selection (or program and provider selection in New Jersey).

• Notification of case manager of enrollment/disenrollment.

• Disenrollment and provider changes.

• Grievances.

Consumer information and education

The public assistance eligibility system and its data files play a critical role in identification of and communication with consumers in all sites. Most programs have personnel assigned to local welfare offices; these personnel describe the program and selection options available to eligible individuals, including using written and audio-visual materials. Only New Jersey has permitted this function to be carried out by provider-based personnel. Other sites permit some dissemination of provider-developed and program-approved promotional materials to aid in enrollee recruitment. Although this education includes an orientation to key features of the demonstration, most programs and providers consider this to be only the beginning of the learning process. This process is meant to give enrollees an understanding of the implications of limited choice and managed care.

Provider selection

All of the demonstration programs are mandatory for targeted eligible groups, with the exception of New Jersey, which has a voluntary enrollment program. Each demonstration does permit and, in fact, requires selection of the participating plan or provider from whom the individual will receive services. Despite this selection opportunity, a high percentage of individuals fail to exercise it. Surveys in Monroe County, for example, suggest that only about two in three recipients make their selection themselves. When no selection is made, various forms of automatic assignment are used. This can produce other problems: In Missouri, it has been surmised that auto-assigned enrollees have higher out-of-plan use rates than self-assigned enrollees; in New York high rates of provider switching among auto-assigned enrollees have led some providers to develop their own schemes of transferring capitation payments in order to reconcile accounts.

Provider notification

Once selection of a provider is completed, this information must be communicated on a timely basis to the responsible plan or provider. Delays in this process, which were common, if not pervasive, in the first year of operation, are problematic for the program, confusing for patients, and costly for providers. Reviews conducted in Monterey after termination noted that as much as $1.5 million dollars in services may have been provided to individuals not appropriately enrolled with the Initiative; thus, the Initiative was not eligible to receive capitation payments from the State for them. The difficulties initially noted in this area have been solved in most sites, although exceptions continue to occur, especially for the more recently implemented programs.

Disenrollment and provider changes

The guarantee of 6-month eligibility in the demonstrations has greatly simplified the disenrollment problem, though disruptions still occur at the end of the guaranteed eligibility period. Another area of considerable concern has been the disruption of patient-provider relationships for individuals whose on-going providers are unwilling or unqualified to become case managers. This concern has been most commonly noted in Santa Barbara, but has arisen elsewhere, especially for the chronically ill and disabled (often supplemental security income-eligible) who have had long-standing provider relationships. This issue has resulted in some program critics and supporters questioning whether case management is appropriate for this class of individuals. In New York, for example, these people are given the opportunity to opt out of the demonstration.

Grievances

All programs provide grievance systems for enrollees to register formally concerns, problems, and complaints about any aspect of the program. The number of grievances have been relatively limited considering the potentially disruptive nature of the demonstrations and the relative generosity of the traditional Medicaid programs in which recipients were previously enrolled. Although most sites are systematically reviewing the nature of grievances to assess overall trends, findings have not been notable. It does appear that as the availability of personnel to handle grievances increase the number of grievances being filed also increases.

Provider participation

The critical issue of provider participation can be explored by looking at three general dimensions: the provider environment, recruitment, and attitudes of participating providers.

Provider environment

In view of the historically low rates of physician participation in Medicaid and the dissatisfaction expressed by many of those who do participate, it was difficult to anticipate how the provider community would respond to these demonstration programs. Characteristically, responses have varied across the demonstration sites, suggesting the importance of local medical service market conditions. The status of the State Medicaid programs, including fiscal crises with anticipated or actual program and payment reductions, has also been widely recognized as having fostered a climate for change, i.e., program reform.

The flexibility of the demonstration programs to involve provider types, which traditionally have not participated in Medicaid, has expanded the options available. However, it has been common for some commercial prepayment plans, like HMO's, to express hesitancy about serving the Medicaid eligible for the first time. For other providers, such as neighborhood health centers, the opportunity to gain experience with prepayment has been welcomed, though with some apprehension. This apprehension is attributed to providers' limited financial reserves to absorb adverse consequences and their lack of knowledge about managing risk. For still other large institutional, often teaching, providers with major commitments to care for the indigent and Medicaid populations, participation was inevitable even if they chose to participate “passively,” i.e., by making few administrative, staffing, and procedural changes in response to program incentives.

Virtually all the demonstrations appear to have benefited from the emerging competitive environment among providers. Characteristics of this environment include: hospital occupancies at unprecedented low levels, a growing surplus of physicians, and growth in group practices, health maintenance organizations (HMO's), and preferred provider organizations (PPO's). Given this environment, providers have shown interest in participation based on the following reasons:

• Opportunity to solidify or expand market share.

• Potential to earn higher incomes from Medicaid patients under the demonstrations than fee-for-service, unmanaged care (by improved control over enrollee utilization).

• Chance to gain the benefits of more timely and predictable cash flow from prepayment.

• Fulfill an expected role for public institutions with large Medicaid constituencies such as municipal hospitals.

This competitive climate is likely to continue for the duration of the demonstrations and beyond.

Provider recruitment

During the development of demonstration programs, most programs, initially, expressed their intent to try to bring into the demonstrations providers who had not previously been major participants with the Medicaid program. These programs were trying to integrate the Medicaid population with mainstream providers and to assure that participating providers could give the desired quality and continuity of care. Some demonstration programs report progress in this direction, although they are more likely to attract traditional Medicaid providers, such as public hospitals and health centers. The recruitment of conventional prepayment organizations has been hampered by a number of factors. Program design features and capitation payment arrangements have effectively excluded HMO's in New Jersey; low rates have discouraged participation in Florida; and general uncertainty about the viability of serving the intermittently eligible Medicaid population has surfaced in a number of programs. As a result of these factors, HMO participation has occurred only in Minnesota (five HMO's), Missouri (a single independent practice association), and New York (a single network-model HMO).

Competition among providers to enroll eligibles has been limited, somewhat at variance with the avowed aim of these competition demonstrations. The reasons for this appear to be related to the following kinds of concerns among providers:

• Is prepayment appropriate for the Medicaid population?

• Do the State or other public agencies have the wherewithal to design, implement, and manage effectively such complex programs?

• Are case-management responsibilities compatible with the primary care provider's other functions and duties?

• Are risk sharing and opportunities for gain appropriately balanced i.e., are rates and methods of payment fair and adequate?

Until these questions are answered so as to allay provider concerns, provider recruitment and, thus, program implementation are delayed. Providers' full potential to compete for more Medicaid recipients can only be realized after these problems are substantially resolved.

It is for this reason that fostering provider competition has emerged as a secondary goal to getting programs implemented. In order to allow the program to gain momentum and credibility, a number of demonstration programs have chosen to negotiate intensively with only a few providers rather than awaiting broader provider participation. The benefits of a spectrum of providers are apparent in the more mature programs, which have enabled providers to learn, initially, that the program is viable; then providers are able to examine their experience to determine if expansion in enrollment is appropriate for them. For example, in Missouri, some providers are now beginning to plan marketing initiatives to expand their enrollments by attracting recipients from their competitors assuming, as discussed earlier, capitation rates remain acceptable.

Participating provider attitudes

The first year studies described how many providers initially reacted to their program responsibilities. To a certain extent, the near universal difficulties with management information systems (MIS's), including the absence of such key program elements as prior authorization procedures, dominated their experiences and attitudes. The second year has seen much improvement in this area and provider attitudes seem to have improved accordingly.

Some of this adjustment must be acknowledged as acclimatization to prepayment for those providers with little or no previous experience with it. This has meant the development of budgeting and other financial systems as well as, in some cases, case management procedures. In addition, making the gatekeeper role an explicit responsibility, in those plans using it, has likewise proven challenging and created a whole new range of issues in inter-provider relationships between primary care physicians, specialists and providers of institutional care. For the hospitals that have chosen to participate as prepaid health plans, varying responses have been noted: Some, like the Monroe County area hospitals, have developed extensive in-house case management systems, and others, as in Missouri, have largely continued providing services as usual. A particularly sensitive issue to be addressed when program outcomes are analyzed, will be whether such providers should be permitted to continue as participating plans if greater cost savings can be achieved without them.

Providers report that they need time to understand and appreciate the subtleties and complexities of case management. Enrollees need learning time as well, especially concerning the lock-in (limitation on choice) aspects of the program. Coordinating the delivery of services takes efforts providers may not have been previously expending and requires development of formal, continuous 24–hour coverage, as well as referral and treatment authorization systems that take time to establish. Programs also have to devise strategies to curb out-of-plan use, including deciding whether to pay other providers for unauthorized care. Interestingly, one of the most irritating aspects of the transition has been the requirement that pseudo (dummy) claims be submitted for prepaid care to enable the demonstrations to be evaluated. For some prepayment organizations such as in Minnesota, preparation of these types of claims is a new responsibility for which additional staff are required. However, most providers report high levels of satisfaction with the efforts of State and Initiative personnel to accommodate their concerns and respond to their problems.

Provider payment and risk sharing

Among the most difficult and critical features in designing the demonstrations has been the complex configurations in the multitiered risk-sharing arrangements developed across the programs. These arrangements, in effect, manifest the assumptions about the kinds of incentives needed to make the demonstrations successful. To illustrate this it is useful to enumerate the levels or tiers around which their payment systems are organized. The tiers include the following:

• State Medicaid agency.

• Risk-assuming intermediary (where applicable).

• Prepaid health plan or organization (where applicable).

• Primary care case manager (where applicable).

• Other providers.

In Figure 2, an overview of these tiers is provided across the demonstrations, illustrating how risk and responsibilities are distributed across the tiers.

Figure 2. Structural relationships of tiers in the demonstration programs.

The State Medicaid agency either directly contracts with providers or engages a risk-assuming intermediary as in the California and New York programs. In these programs, the State delegates the administration of the program to such an intermediary and the State also sets a capitation rate to enable it to acquire covered services for enrollees. This is customarily a discounted rate (usually 95 percent), based on the historical fee-for-service equivalent payment. The principal advantage of this system is that the State can guarantee itself savings, and the intermediary has an incentive to acquire services for enrollees at the lowest available cost.

The next tier, the prepaid health plan or organization, may be engaged by the State directly or by the intermediary to assume responsibility for providing covered services. In the State-administered programs, such as Missouri and Minnesota, these plans are capitated to place the prepaid health plan or organization at financial risk. Missouri does permit an exception to this, as noted earlier, with the physician sponsor program, which is not capitated but is paid fee-for-service with a case-management fee. In New York, where MediCap is a capitated risk-assuming intermediary, the prepaid health plan is a network-model HMO and is also capitated, functioning like a secondary intermediary.

Significant variation among the programs is found at the level of the PCCM. In some demonstration programs, case management is an explicit component. In others, case management is not a uniform feature, though prepaid plans may elect to use it as a cost and utilization control technique. In Santa Barbara, the intermediary requires participating providers to be case managers and pays the PCCM's on a capitated basis for primary care services; in Monterey, the method of payment was fee-for-service with a case-management fee, similar to the Missouri Physician Sponsors Plan. In Itasca (MN), the county is capitated and the PCCM's are paid fee-for-service with both the county and the PCCM's sharing in surpluses or deficits. In the other Missouri provider arrangements and in Hennepin and Dakota programs, the prepaid plans IPA's, HMO's, neighborhood health centers, and hospitals, may elect one of the following:

• Not to have individually responsible case managers.

• To contract with and capitate PCCM's.

• To employ a PCCM physician and pay a salary.

In New York and New Jersey participating providers are expected to adopt the case-management approach. These variations in commitment to and employment of the case-management concept typify differing assumptions about its expected usefulness in containing costs and improving access.

The final tier relates to risk sharing for nonprimary care providers, including inpatient care, medical specialists, or nonphysician providers. The demonstrations are experimenting with a number of arrangements, ranging from separating this entirely from the primary care payment systems to putting the prepaid plan or PCCM at full risk for all care. It is at this tier that the treatment and referral authorization systems become highly important; the systems are closely linked with how financial risk for nonprimary care services is apportioned among the various tiers.

At all of these tiers variations can be found reflecting the exploratory and adaptive nature of the risk-sharing process. The variations may be attributed to the assumptions of program developers about the effectiveness of various risk-sharing-arrangements. Further, the variations suggest that some programs attempt to be highly explicit about how participating organizations and individuals are to achieve cost savings; others leave these decisions to the managerial discretion of the plans and providers.

Rate setting

Equally complex, and perhaps more controversial, are the rate setting methods employed across the demonstrations. Although still emerging during the initial stages of implementation, rate setting has now arrived at center stage, especially for the mature programs assessing the long-term possibilities of case management. Some program managers and providers assert rate setting will be the single most important issue in determining program viability.

Most demonstrations began operations with the goal of outperforming, i.e., having costs lower than, the existing fee-for-service equivalent costs for eligible care, generally on the order of 5 percent. Relying on consultants and other resources, the States arrived at actuarially determined costs of covered care for various rating categories, with some sites using as few as 2 categories (AFDC adults and AFDC children) and others use more than 70, as in Minnesota. These costs were then trended forward; adjusted for geographic differences; and deductions from costs were made for various reinsurance or stop-loss arrangements before final distribution of the costs to the appropriate program funds for provider disbursement. Some programs, such as Missouri, have established risk pools for special groups such as newborns with major medical problems; these programs have funded such groups with mandatory deductions from the capitation rates of all programs.

Even assuming a stable base, numerous complexities soon began to surface. They took on considerable importance given the tenuous nature of relationships with skeptical providers. Questions about the composition, homogeneity, and number of rating categories emerged. Trending factors were challenged. The use of local recipient experience rather than statewide experience was challenged, especially if the number of local Medicaid eligibles was small. Documentation to support the methodologies was also inadequate, inconsistent, or absent. The deductions made for funding reinsurance and stop-loss coverage, as well as the computation formulae used, have also been disputed in some demonstrations. Delays in getting rates approved at the State or Federal level occurred.

A more confounding problem arose when it became apparent that because of other program reforms and larger scale changes in health services, the fee-for-service base was not stable and evidently declining in a number of programs. Thus, when second year rates were computed some were found to be lower than first year rates; the differences were substantial amounts in some areas, such as New Jersey. Some observers contend that these pressures are just what are needed to compel PCCM's to manage even more effectively to justify the programs' existence; others are concerned that these pressures may have prevented the demonstrations from being able to test adequately the strategies being implemented. In either case, this issue has the potential to inhibit severely physician recruitment and to destabilize provider relations and participation. Consequently, intensive discussions and negotiations to address this program are under way in a number of program sites. Within the evaluation of the demonstrations, the rate setting methodologies and processes are being extensively reviewed.

Management information systems

The severe management information system (MIS) problems of the first year are being solved at most of the program sites. For some programs, this has meant refinement and redesign, establishing supplemental systems or replacing contractors who failed to produce usable systems on a timely basis. Despite this progress, problems are still apparent; they are exacerbated by growing provider interest in more sophisticated systems, which will enhance providers' ability to carry out their responsibilities in the demonstration programs.

It is important to note that while having an MIS is not an assurance of an effective program, its absence has profound negative consequences in such areas as the following:

• Program operation and assessment.

• Eligibility and enrollment linkages.

• Provider participation and payment.

• Financial monitoring.

• Utilization review and management.

• Quality assurance.

In addition to having system components to support each of these areas, the coordination and report generation from them must be precise and timely to facilitate such activities as prior authorization of specialty services or preadmission certification, which some plans and PCCM's are implementing.

The interrelationship between incentives and provider behavior becomes apparent when looking at such areas as utilization monitoring. Where plans and case managers are at financial risk for specialty care, they wish to be positioned to be aware of and, perhaps deny, unauthorized out-of-plan care. When such problems appear extensive or persistent, the PCCM may then implement more stringent authorization measures. In Missouri, plans have had to decide whether to reimburse other providers for unauthorized out-of-plan use. In some cases these other providers may be competing prepaid health plans that are well aware of the demonstration program and its lock-in provisions for enrollees. When MIS reports are unavailable or unusable, it is not possible to monitor care closely. For example, the reports of specialty use in one program given to the PCCM list the specialist only by Medicaid provider number rather than by name; this makes it difficult for the PCCM to identify and resolve unauthorized use problems.

Quality assurance

The second year of demonstration programs has witnessed increased attention to quality assurance as well as utilization review. For a program using prepayment and limitations on choice of provider, and an overarching goal of cost containment, concern about under-utilization is generally regarded as the principal quality of care concern. Stated differently, the service use to be reduced by the demonstration programs is intended to be only unnecessary care. Because of this focus, much of what has been cited as quality assurance activities are largely utilization review issues.

However, some more typically quality assurance activities are now occuring at various demonstration sites:

• Employment of clinical personnel at the State or risk-assuming intermediary tiers to oversee or conduct quality assurance efforts.

• Monitoring of 24 hour availability of the PCCM.

• On-site medical record audits.

• Operationalizing of quality assurance plans and committees by providers.

• Small scale treatment outcome or sentinel event studies across providers.

• Development of clinical management protocols for selected high prevalence conditions.

Notwithstanding these examples, the programs uniformly cite quality assurance as an area to which they will devote additional attention and resources in the next year.

Managerial concerns

The final issue which incorporates many elements of those presented earlier, is program management. These programs have severely tested the developing agencies' abilities. Agencies have had to take programs from conceptualization to full implementation in highly compressed time periods. Because most Medicaid agencies have neither the organizational slack nor many of the requisite technical skills in-house, reliance on outside consultants has been extensive. At best, this added another layer of administrative complexity and, at worst, it has provided the basis for serious conflict, especially when consultant nonperformance has become an issue.

A core group of committed staff has proven critical in certain program sites. Other program sites have experienced substantial turnover but have still been successful, suggesting that factors beyond permanency of personnel play a role. The tensions between delegating and centralizing functions, as noted in New Jersey and elsewhere, have also been played out differently, assuming that some minimum, adequate number of personnel are engaged in the key program operations. Provider perceptions are also important. Providers have reported how disconcerting it can be to have to deal with a stream of unfamiliar and continually changing personnel.

The evidence on the advantages and disadvantages of the risk-assuming intermediary tier versus direct contracting between the State and prepaid health plans and providers is mixed and inconclusive. The risk-assuming intermediary can link and tailor a program to a local market, but it also adds another party to the complex round of negotiations required to get a demonstration program started. In addition, as some critics suggest for Monterey, the strained State-Initiative relationship may be a contributing factor in the ultimate demise of a program. For New York, some have questioned the role of MediCap when only a single provider network is participating. Given this situation, MediCap's position would seem to be duplicative, at least until other plans are recruited.

Key issues emerging in year three

As previously noted, the continuing programs remain at various points of development and maturation as many enter what is expected to be the final year for them, unless extensions are granted. A number of important developments are expected:

• Transition to permanent status for certain programs.

• Continued transformations in local health service markets.

• Increased evidence of competition among providers for enrollees.

• Rate setting to become more contentious.

• Quality assurance programs to become more prevalent and stringent.

• Case management to be better understood by providers and enrollees.

• Appropriateness of case management to be challenged for selected eligibles.

These issues are now briefly described.

Transition

Santa Barbara, Missouri, and New Jersey have requested waivers (section 1915b) from HCFA to continue their demonstration programs when the demonstration funding expires in 1986. Such waivers are required because a number of program features represent exceptions to conventional Medicaid program requirements and thus must be specifically exempted by the waiver process. The Santa Barbara and Missouri programs are likely to be approved. Few significant changes are expected because both report that their own cost analyses suggest positive financial results, a requirement for granting the waiver. As discussed previously, rate setting will be an issue of major importance in both of the programs. In California, it will be an issue because of the shrinking fee-for-service base and the administrative cost dispute with the State; in Missouri, it will be an issue because of the program's expressed interest in going to provider-specific capitation rates. More competition is expected among providers if rate setting is perceived as satisfactory. The New Jersey waiver request is currently under review.

Transformations in local health service markets

As hospital occupancies continue to decline, alternative delivery system enrollment will continue to grow, and competition will grow more fierce. The capitated demonstrations are likely to receive at least indirect support from these larger market forces, especially as fee-for-service payment becomes the exception rather than the rule, as it appears to have in such places as Minneapolis. To a limited extent, the demonstrations have stimulated interest in prepayment among providers, like the neighborhood health centers in Missouri; the demonstrations have given providers much needed experience with prepayment. Despite these changes, it is not yet clear if program designs or recruitment strategies will succeed in bringing more previously non-Medicaid providers into participation. These strategies may simply convert traditional providers to prepayment.

Competition among providers for enrollees

Even if few additional providers enter the demonstrations or their successor programs, it is expected that where the program proves creditable and feasible, economies of scale will be pursued. It will be of interest to see if, given the nature of the mandatory basic service coverage of the programs, some providers attempt to add optional services to attract enrollees. Another alternative would be more intensive media-related publicity efforts, which have not proven particularly effective in affecting initial plan/provider choice. This competition also presumes the maintenance of capitation and other rates that are acceptable.

Rate serving conflicts

In order to assure provider participation the programs are required to pay rates that are perceived by providers as adequate. Florida's demonstration was unable to recruit prepaid health plans because it offered rates discounted from what were already among the lowest Medicaid fee-for-service rates in the country. This is one of the principal lessons of the Florida demonstration failure. For program managers and providers, the negotiation and retention of adequate rates is likely to be a source of severe conflict, especially if the program has proved to be feasible and profitable. The issue of adequate rates has significant political and equity overtones, as well as technical ones. These overtones suggest that easy solutions will not be found despite the clear aim of cost containment.

Quality assurance

For programs that have demonstrated that they can be implemented and cost savings can be achieved, the next questions which inevitably arise are how were the savings attained and what may have been given up. This issue is likely to intensify interest in finding out if the reductions in service cost, use, or substitution effects of less expensive for more expensive care, have had adverse health consequences.

Understanding case management

Despite assertions by many primary care providers that they have always been case managers, the embodiments of case management found in many of the demonstration programs has taken time to learn and understand—both for providers and enrollees. This is important to note for two reasons: first, learning effects are more likely to be apparent in provider and Medicaid recipient behavior as more as time passes; and, second, gaming of the system is likely to increase as sophistication grows. This latter point may apply both to the recipient who realizes that the emergency room is unlikely to turn away an insistent but unauthorized patient, and to the provider who may try to encourage high-risk individuals to enroll elsewhere. The key issue is that case management, like prepayment enrollment, provides an acculturation experience that will take time to absorb.

Appropriateness of case management

Some evidence has already emerged that primary care case management may be inappropriate for certain patients with long-standing provider relationships for chronic conditions. To the extent these providers are not candidates for becoming primary care case managers, disruptions and discontinuity may result. These patients and others with pre-existing conditions also present problems of adverse selection for providers with whom they do enroll, sometimes requiring setting up complicated risk pools for such circumstances. It is likely that other programs, particularly those that cover the disabled populations as well as AFDC and SSI eligibles, will exempt these patients from the conventional case-management program, as has been done in Monroe County, or will devise some alternative program for them.

Conclusion

Significant progress occurred in most of the demonstration sites during this period of time. The problems addressed by most programs have been ones of development and implementation rather than of design and consensus building which marked the first year. Much more has been learned about the feasibility and difficulty of making these programs work; in two cases (Santa Barbara and Missouri) it has become apparent that the programs will continue after the demonstration has been concluded. However, the answers to many other questions are inconclusive, and the long-term fate of the other programs, including their cost-containment strategies, are still unknown.