Evaluation of Arizona Health Care Cost Containment System, 1984-85

Abstract

In this article, we describe the evaluation of the Arizona Health Care Cost Containment System (AHCCCS), Arizona's alternative to the acute care portion of Medicaid. We provide an assessment of implementation of the program's innovative features during its second 18 months of operation, from April 1984 through September 1985. Included in the evaluation are assessments of the administration of the program, provider relations, eligibility, enrollment and marketing, information systems, quality assurance and member satisfaction activities, the relationship of the county governments to AHCCCS, the competitive bidding process, and the plans and their financial status.

Introduction

The Arizona Health Care Cost Containment System (AHCCCS) is an innovative system for providing acute medical care services to the indigent population in Arizona, the only State without a traditional Medicaid program. The Arizona State government received Federal funding for AHCCCS as a demonstration project of the Health Care Financing Administration (HCFA).

AHCCCS differs from the indigent health care programs of other States in several respects. It covers acute medical care services similar to those covered by a traditional Medicaid program, but it does not cover any long-term care services. It selects its providers through a competitive bid process; these providers are reimbursed under a prepaid capitation system. The State government itself is also reimbursed on a prepaid capitation basis by the Federal Government. Under the program, beneficiaries are assigned to a particular “gatekeeper,” who manages their medical care, and beneficiaries are required to pay small copayments for services they receive. The original legislation required that most of the program's administrative functions be contracted to a private administrator. The State government has now taken over this function, although it has retained the option to contract out specific functions. In addition, the original legislation called for the program to include private, State, and county employees in addition to the indigent. To date, these groups have not been included. The county governments determine eligibility for medically indigent (MI) and medically needy (MN) beneficiaries and continue to provide long-term care services and other services they had previously provided that are not covered by AHCCCS.

The AHCCCS innovations potentially can be replicated in other geographic areas and for programs other than Medicaid. Consequently, it is important to determine how well the program has worked. In particular, it is necessary to determine whether the program provides access to high-quality care at lower cost than do conventional Medicaid programs.

Before AHCCCS was instituted, indigent-care programs in Arizona were run by the individual county governments. The eligibility criteria, as well as the services to be covered, were determined in each county and differed from county to county. Over the years, legislation to participate in the Federal Medicaid program was routinely introduced and defeated or passed and not funded. In 1980, when legislation was passed limiting local property taxes in Arizona, county revenue was no longer adequate to cover the rising cost of indigent health care, and the counties needed to find a way to get Federal and State support. However, many Arizona legislators continued to be skeptical about participating in the Federal Medicaid program because they were concerned about program costs. In 1981, negotiations between HCFA and Arizona legislators concluded with an agreement on a 3-year demonstration project, which began in October 1982. This demonstration project has been extended through September 1988, and negotiations are currently under way to extend Federal support for the program for an additional 5 years if long-term care is added.

Eligibility for AHCCCS includes all groups categorically eligible for Medicaid—Aid to Families with Dependent Children (AFDC) and Supplemental Security Income (SSI) recipients—and the medically indigent and medically needy. AFDC and SSI are joint Federal-State programs for which the State receives Federal matching funds (Federal Medicaid assistance percentage). People eligible for these programs are considered categorically eligible for AHCCCS. MI and MN beneficiaries are poor individuals who meet Arizona requirements for eligibility for AHCCCS services but who are not categorically eligible. The State does not receive Federal matching funds for MI or MN beneficiaries. As of May 1984, of the approximately 190,000 beneficiaries eligible for AHCCCS, 44 percent were AFDC beneficiaries, 19 percent were SSI beneficiaries, and 38 percent were MI or MN beneficiaries. In September 1985, 160,369 people were eligible for AHCCCS. Of these, 50 percent were AFDC beneficiaries, 21 percent were SSI beneficiaries, and 29 percent were MI or MN beneficiaries.

Benefits covered include most acute care services: hospital, physician, laboratory, X-ray, medical supplies, pharmacy, and emergency services. Skilled nursing facility and home health services are not included in AHCCCS benefits.

AHCCCS is jointly funded by the Federal Government, the Arizona county governments, and the State government of Arizona. The Federal Government pays a capitation payment for federally qualified eligibles, the counties provide a percentage of their pre-AHCCCS indigent care budgets, and the State makes up the difference. From July 1982 through July 1985, the Federal Government contributed 27 percent of the $585 million in revenue, the counties 30 percent, and the State 39 percent. The remainder was contributed by earned interest and third-party payments.

About two-thirds of the expenditures of the program from July 1982 through June 1985 were for capitation payments to the plans, and a little more than one-fifth were for fee-for-service payments to providers. Smaller amounts were for program administration (6 percent), reinsurance (2 percent), Medicare Part B premiums (1 percent), and services for crippled children (1 percent).

The evaluation of this demonstration is being conducted under contract to HCFA by SRI International in Menlo Park, California. The evaluation team also includes Actuarial Research Corporation in Annandale, Virginia, and Research Triangle Institute in Research Triangle Park, North Carolina. The evaluation is divided into two main parts: assessment of AHCCCS program outcomes and an explanation of the implementation and operation of specified AHCCCS program features.

In the first 18 months of the program, the evaluation was focused on the program's implementation and operation, in which there were substantial problems. AHCCCS was implemented quickly, and the administrative functions were contracted out without sufficient thought to what was being contracted for or how to monitor performance. Eligibility determination and enrollment created difficulties for the counties, the administrator, and the State. As a result, a large number of fee-for-service claims had to be paid for beneficiaries requiring health care during the time between eligibility determination and enrollment. Plans did not know who their members were. MCAUTO Systems Group, Inc. (MSGI), the plan administrator, did not know who was enrolled, and the counties had difficulty making eligibility determinations in the allocated time. Plans bid, were awarded contracts, and provided services to beneficiaries, but the lack of data on the number and kinds of patient encounters made it impossible to know whether more beneficiaries were being served than under the previous indigent care system. Emphasis on day-to-day problems caused a lack of attention to monitoring the collection of data on cost and encounters. In addition, little attention was given to the development of quality assurance or patient satisfaction procedures. Private-sector involvement was postponed in part because of the projected high bid and administrative costs and in part because of the poor public perceptions of the program.

On the other hand, the program was able to attract a large number of full-service bidders because provider acceptance was not a problem. Prepaid plans were formed where few had previously existed. These plans developed internal cost-efficiency measures in an effort to contain their costs. This prepaid-plan formation not only had an impact on the availability of medical care to the indigent but also promoted the development of prepaid options for the private sector.

By the end of the first 18 months of program operation, statewide eligibility criteria had increased the number of persons enrolled in the program to more than had been enrolled under the previous county system, and quality of care did not appear to be a problem. Reviews of the health plans by the Accreditation Association for Ambulatory Health Care found that “the quality of the medical care provided to AHCCCS patients appears to be at least equivalent to the care rendered by AHCCCS providers to their private non-AHCCCS patients” (Moen, 1985). In addition, the program had an impact on the entire health care delivery system in Arizona by creating a more cost-sensitive environment.

In this article, we examine program implementation and operation issues in the second 18 months and focus on the administration of the program, provider relations, enrollment, information systems, quality assurance, relations with the Arizona counties, the bid process, and the plans and their financial status. Other evaluation reports to be produced in the next year will be focused on the cost of the program and on beneficiary access to and satisfaction with the program. Later reports will be focused on the quality of care and utilization of medical care services by program beneficiaries.

Administration of program

The administration of AHCCCS during the second 18 months of the program was significantly different from the administration during the first 18 months. The original AHCCCS legislation required that major parts of the program's administration be contracted to a private administrator. MSGI was selected to be the private administrator and served in that capacity for 18 months.

During the first 18 months, AHCCCS suffered from two major administrative problems: divided responsibility between the private administrator and the State government and inadequate staffing for essential responsibilities. As a result, the private administrator experienced difficulties in carrying out its functions in critical administrative areas, including eligibility and enrollment, claims and encounter data processing, provider relations, financial monitoring, and quality assurance. On March 15, 1984, the State government took over the private administrator's functions, following a contract dispute with MSGI.

After the State government took over responsibility, two major organizational changes brought about a new structure for AHCCCS. First, the legislature created the AHCCCS administration, separate from the Arizona Department of Health, to run the program on a day-to-day basis. Second, the new State AHCCCS administration was given authority by the legislature to run the entire program, thus ending the dual system of policymaking by the State AHCCCS Division of the Department of Health and administration by MSGI that had existed in the first 18 months.

The State AHCCCS administration created a more unified organizational structure and added significant staff to key areas that had been understaffed during the first 18 months: quality assurance, financial monitoring, provider relations, information systems, and enrollment.

In the new organizational structure, key program functions were organized into four major operating divisions (Provider and Member Services, Information Systems, Financial Management, and Business Systems) and four major staff offices (Medical Director, Audits and Compliance, Grievance and Appeals, and Intergovernmental Relations and Planning). The total staff size of AHCCCS by September 1985 was 301, and total administrative expenditures in State fiscal year 1984-85 were $15 million. Administration accounted for 6 percent of total program expenditures during State fiscal year 1984-85, compared with 8 percent during State fiscal year 1982-83.

Although major administrative responsibilities are no longer contracted out to a private firm, the State AHCCCS administration has used outside contractors for specific purposes. Financial reviews of the prepaid plans and medical audits have been contracted. Private contractors have also been used to help design the new information system. The main change from the first 18 months, therefore, is that AHCCCS has chosen to use private contractors when necessary and has not chosen to contract out major responsibilities.

The AHCCCS administration gave special attention to ensuring the financial viability of plans and performing quality assurance activities, and work began on developing a new information system. From a financial review of the plans conducted during 1984, the Director of AHCCCS was able to determine that several prepaid plans were experiencing problems of financial solvency. The Director terminated the contract with two plans and encouraged a third plan to reorganize.

Significant progress has been made in the area of quality assurance. The Office of the Medical Director has given major attention to conducting medical audits of the plans and providing technical assistance to them when corrective actions are necessary. AHCCCS also increased its activity in resolving grievances and appeals, eliminating a substantial backlog built up during the first 18 months.

With respect to information systems, the AHCCCS administration has begun an effort to design a new system that will be specifically tailored to the prepaid environment. While this development is under way, major problems remain with the implementation of the current encounter data processing system. The AHCCCS administration has given attention to this problem, including the creation of a special unit, but it had not, as of October 1985, been able to collect and process encounter data from the prepaid plans on a regular and consistent basis.

Overall, AHCCCS made substantial progress in stabilizing the administration of the program during the second 18 months by addressing many of the critical problems encountered during the first 18 months. By the end of the third year of AHCCCS, most of the essential elements for administering the program were in place. The major remaining area of administrative difficulty was implementing an effective information system. The administration hopes to address this difficulty through the development of a new information system specifically designed for the unique features of AHCCCS.

Provider relations

In a prepaid program, the State government has a responsibility to monitor and provide technical assistance to the contracted health plans. This provider management function is necessary to ensure appropriate plan performance in the areas of quality of care, delivery of proposed services, utilization, contract compliance, financial condition, marketing activities, facilities, and member rights and patient satisfaction.

In its proposal to become the AHCCCS private administrator, MSGI anticipated a comprehensive program of plan management, including various activities in the areas of technical assistance, utilization review, quality assurance, performance monitoring, contract compliance, and overall program communications.

Unfortunately, during the early months of program operations, the activities in this area fell considerably short of expectations. Because of attention required to address early startup problems and because of the lack of sufficient resources, the administrator's provider management function initially was focused on activities essential to keeping the program running on a day-to-day basis. Resolving daily operational problems, especially those related to enrollment difficulties and provider payment problems, consumed most of the staff's energy. The provider management function became one of reacting rather than the initiating role that had been anticipated.

Provider management activities increased substantially following the transition to State government administration. A number of factors were important in this change. First, the responsibilities for provider management were defined and assigned to specific organizational units. Second, resources allocated to provider management activities throughout the organization were substantially increased. Third, the AHCCCS staff gained experience, which improved their administration of this function. Finally, and perhaps most importantly, the new AHCCCS leadership voiced a strong commitment to a vigorous provider management function.

Most of the organizational units within the AHCCCS administration have some area of responsibility for provider management. The major provider management activities in place at the end of the third program year include overall coordination of plan management activities by six plan managers assigned to the individual plans; monthly compliance assessments of various aspects of plan operations; annual financial audit of plans and quarterly review of financial reports; fraud and abuse investigation; annual medical audit of plans; monitoring of plans' quality assurance activities; reviews of plans' utilization patterns; periodic seminars and meetings of the plans' medical directors; monitoring of plans' submission of encounter data and provision of technical assistance when needed; assistance in billing procedures for fee-for-service claims; workshops and training sessions on a variety of topics; and distribution of newsletters, policy memos, and manuals to keep plans and providers up to date on AHCCCS policies and procedures.

Eligibility, enrollment, and marketing

In a prepaid program, eligibility determination, enrollment, and marketing are critical areas of operation that must be carefully coordinated and closely linked. Information systems must be in place to track member eligibility and enrollment status, and consistent, up-to-date enrollment information must be available for membership verification. Successful enrollment of members in health plans is crucial because it marks the beginning of plan responsibility for delivery of care to the members and forms the basis for capitation payments to the plans.

In AHCCCS, eligibility and enrollment responsibilities are shared among the county governments, AHCCCS staff, the State Department of Economic Security, and the Social Security Administration. This division of responsibility was a cause of significant difficulties in the first 18 months of the program. To a large extent, these difficulties related to differences in enrollment status as perceived by the members, the plans and providers, AHCCCS, and the county governments and agencies determining eligibility. Because of timelags inherent in the eligibility and enrollment process and the division of responsibility among several parties, the information about an individual member's enrollment status often was not consistent from one location to another. This inconsistency created confusion regarding the member's true status and was an obstacle to the delivery of and payment for services to the member.

Most of the early enrollment problems were resolved in the second 18 months of the program. This was accomplished through the implementation of improved information systems linking the various parties and through simplification of the enrollment process itself.

Eligibility determination and enrollment of MI-MN beneficiaries created special problems, which resulted partly from the fact that these beneficiaries often become eligible when they are sick and, therefore, may need services before they are enrolled in a capitated plan. As a result, the MI-MN population created a large fee-for-service liability for AHCCCS. Problems also exist because the county governments are responsible for determining MI-MN eligibility. The interests of the counties and AHCCCS may conflict in this area because increasing enrollment in AHCCCS minimizes the cost of ongoing county indigent care responsibilities but increases the cost of AHCCCS.

Both of these problems were addressed during the second 18 months of AHCCCS. AHCCCS's fee-for-service liability was reduced by automatically enrolling MI-MN beneficiaries in a plan immediately upon eligibility determination. This automatic assignment process also ensured that members, plans, county governments, and AHCCCS all had consistent information regarding the individual's enrollment status. However, this improvement was made at the expense of reduced freedom of choice for MI-MN beneficiaries. In addition, automatic assignment can have adverse financial effects for the plans because they may become responsible for MI-MN members at the time they are using services. To counter these effects, AHCCCS adopted a “deferred liability” policy to reduce the plans' financial liability. More rigorous documentation and verification procedures also were instituted for county MI-MN eligibility determinations. Although AHCCCS believes that these procedures ensure that only people who are eligible are enrolled in AHCCCS, the county governments contend that these new procedures have discouraged people from completing eligibility determination and have excluded persons eligible for MI-MN coverage from AHCCCS.

The scope of permitted plan marketing activities was significantly limited during the second 18 months of AHCCCS, although there had been no major marketing abuses during the first 18 months. There continued to be no significant problems, and plans did not object to the increased restrictions.

Information systems

The primary objectives of the AHCCCS management information system (MIS) are to:

• Issue capitation payments to plans.

• Receive and edit encounter data from capitated providers.

• Receive, edit, and pay fee-for-service claims.

• Maintain records of individual members and providers.

• Produce financial, administrative, and utilization reports to support the management of the program.

An effective management information system is a critical component of any prepaid program. The MIS requirements in a prepaid program are significantly different from those in a traditional fee-for-service program. Some of the key areas of difference concern the processing and maintenance of and access to enrollment information; processing of capitation payments; processing and reporting of encounter information; maintenance of prepaid-plan information and provider affiliations; processing of reinsurance claims; and production of utilization summaries and management reports that are focused on the particular management concerns of a prepaid program.

AHCCCS experienced major difficulties in implementing its management information system. The original approach taken was to adapt an existing Medicaid Management Information System to the unique needs of AHCCCS. This approach was expected to result in the most timely and cost-effective implementation of an MIS. In practice, it fell considerably short of expectations. The differences between the requirements of a prepaid program and those of a traditional fee-for-service environment were found to be significantly greater than originally anticipated. Consequently, implementation of major portions of the MIS was delayed significantly, and the resulting system did not meet many of the needs of the AHCCCS users. Implementation was further hindered by the short time allowed for development and installation, the turnover of key staff, the lack of sufficient human resources, the rapidity of changes in policy and regulation, and the fact that this was the first program in Arizona under Title XIX of the Social Security Act.

At the time of the transition to State government administration in March 1984, the computer system was capable of performing the basic day-to-day transaction processing functions for AHCCCS. However, the MIS could not properly be characterized as an “information system,” especially given the long delays in implementing its reporting subsystems and the difficulty in accessing stored information.

Not surprisingly, the program's MIS needs were an early focus of attention of the new State AHCCCS administration. A key element of this focus was improving the quality of the information available to manage the program; it was critical that the MIS become far more than just a transaction processing system.

The new AHCCCS administration began to address the information system needs in two parallel thrusts: first, to enhance the existing MIS wherever such efforts would be cost effective, given the constraints of that system; second, to explore the possibility of developing a new MIS designed specifically for the unique needs of a prepaid program. AHCCCS has submitted to HCFA an advance planning document outlining its MIS development project, with an expected implementation date of April 1988. HCFA has approved the early phases of that project relating to the development of a general system design.

The encounter data receipt and processing operation of the MIS was severely hampered by the failure of the plans to submit encounter data. In the first 18 months of the program, only a fraction of the expected encounter data were collected. The collection of such data had a low priority for the State government and the submission was a low priority for the plans. In March 1984, when the State government took over the administration of AHCCCS, one of the areas HCFA wished to make a priority of the new administration was the submission of complete and accurate encounter data.

During the next 18 months of the program, a work plan was developed, improvements were made in the AHCCCS encounter data process, and outside contractors were hired to assist in encounter data processing. These efforts resulted in a slow but definite improvement in the amount of data collected and its credibility. However, the plans still submitted less data than would have been expected. Our analysis of the data in the AHCCCS program files as of April 1, 1985, for services received during the first 3 months of the third year of service (October-December 1984) indicated that less than two-thirds of the number of encounters predicted had been submitted to AHCCCS. As of November 1985, encounter data processing was still not operating smoothly, and an effective system of financial sanctions had not been developed. The November 1985 yearly review letter from the HCFA San Francisco Regional Office to AHCCCS pointed out that the State government needs to apply penalties against health plans that do not submit encounter data (O'Connor, 1985).

Overall, there is more responsiveness to encounter data problems within AHCCCS now than at the end of the first 18 months of the program. This change, coupled with an extended effort to train and monitor the plans, has led to an increase in the quantity and quality of encounter data. However, there are still serious questions about the ability of the program to provide the required data and to process those data into a form necessary to manage the program.

Quality assurance and member satisfaction

Quality assurance and member satisfaction procedures are essential to ensuring that adequate care is provided under AHCCCS because prepayment has introduced changes in provider incentives that may affect the quality of care. Both the AHCCCS administration and individual plans have responsibilities for quality assurance and member satisfaction. These areas were largely neglected during the first 18 months of AHCCCS. However, during the second 18 months, activities increased substantially, and quality assurance, in particular, was a priority for AHCCCS. The AHCCCS administration improved its procedures for monitoring quality of care and more strictly enforced provider contract requirements in this area. The annual medical audits conducted by AHCCCS have evolved, and the procedures used have become more sophisticated. Although there was also increased attention in the area of member satisfaction, AHCCCS still lacked an effective system for monitoring the resolution of grievances at the plan and State levels or for directly monitoring member satisfaction through periodic surveys.

Frequent turnover and vacancies in the medical director's position were impediments to the development of quality assurance activities during the first 18 months of AHCCCS. The situation improved in the second 18 months with the appointment of a full-time medical director in June 1984 and creation of the Office of the Medical Director. Since then, the resources committed to planning for the annual medical audits have increased substantially, and, particularly in the third program year, an improved methodology has been used in the audits. Developing skills in quality assurance at the plan level has also been identified as a major goal, and AHCCCS has actively monitored plan implementation of quality assurance programs and has provided technical assistance to the plans in developing their programs. Although the AHCCCS administration has taken important steps in developing its quality assurance program and undertakes more extensive activities than any State Medicaid program, its activities could be further improved by additional refinement of the procedures used in the annual medical audits, strengthening mechanisms for feedback to providers, initiating more comprehensive followup actions to produce improvements in quality of care, developing stronger systems for routine monitoring of care, and defining specific standards for plan quality assurance activities.

The AHCCCS administration is required to conduct annual medical audits of all contracting plans, and each year separate audits have been conducted for regular AHCCCS plans and for plans providing services to those in long-term care institutions. As in the first program year, AHCCCS contracted with outside organizations to conduct the second- and third-year medical audits. The review of regular plans was conducted in each year by the Accreditation Association for Ambulatory Health Care (AAAHC). Each year, the methodology has changed in some respects. In particular, efforts were made in the third year to decrease the subjectivity of the assessment of quality of care by, for the first time, auditing samples of patient records for compliance with specific treatment criteria established for selected diagnoses. Despite these improvements in the audit methodology, it is difficult to draw conclusions about the quality of care provided at the plan or program level because of wide variation among providers as well as among plans. AAAHC did find that, compared with the first year of AHCCCS, when only five plans had any quality assurance procedures, all but three plans had a quality assurance program in the second year, and all plans had such a program in the third year.

Member satisfaction also received more attention during the second 18 months, primarily in the area of grievances and appeals. However, in general, less progress was made in this area than in quality assurance. There was virtually no activity in hearing grievances and appeals from April 1984, when the State government assumed administration of AHCCCS, to December 1984, when the Office of Grievance and Appeals was formed in the AHCCCS administration. During this time, a large backlog of overdue, unresolved cases had developed. Much of this backlog was eliminated by the end of the second 18 months. Most of the grievances, however, were not submitted by plan members and did not relate to member satisfaction. From May 1984 to September 1985, only 53 formal member grievances were submitted to AHCCCS.

Member grievances are also heard at the plan level, and complaints may be handled by the plans or any of a number of offices and divisions within the AHCCCS administration. Although AHCCCS established formal procedures at the plan and program levels for addressing member complaints and grievances, it has not effectively monitored implementation of these procedures by the plans or coordinated activities within the AHCCCS administration. Thus, by the end of the second 18 months, AHCCCS did not have a system to ensure that complaints and grievances were handled according to procedures and resolved in a timely manner or to routinely track the number and types of member complaints and grievances filed.

In addition, AHCCCS did little to monitor member satisfaction directly, for example, through member surveys. In the third year, some member surveys containing questions related to satisfaction (e.g., reasons for switching plans, understanding of grievance procedures, and satisfaction with the resolution of grievances) were fielded, but the response rates to these surveys were so low that no reliable conclusions could be drawn.

County governments: Relationship to AHCCCS

Before the implementation of AHCCCS, the Arizona county governments had sole responsibility for providing health care services to the indigent. Since the creation of AHCCCS, much of this responsibility has shifted to the State level. However, the role of the county governments has not been eliminated altogether. In addition to their responsibility for eligibility determination, the counties relate to AHCCCS and the delivery of indigent health care services in Arizona in three main ways:

• County governments continue to be responsible for providing long-term care services, which are excluded from AHCCCS through waivers of Medicaid program requirements granted by HCFA.

• Counties continue to contribute to the financing of indigent health care services, both those provided under AHCCCS and those they provide directly.

• Counties that operate hospitals continue to act as direct providers of services for both indigents who are eligible for AHCCCS and those who are not.

The exclusion of long-term care from AHCCCS has created problems since the beginning of the program, and problems encountered during the first 18 months continued during the second 18 months. It has proven difficult to coordinate the provision of acute care financed by AHCCCS and long-term care financed by the county governments. The situation is particularly complicated because AHCCCS does cover acute care services provided to eligible people in long-term care facilities. These people are considered a special population covered by AHCCCS and are enrolled in a long-term care, rather than a regular, AHCCCS plan. In counties for which no contract is issued, services are provided on a fee-for-service basis with a payment limit, or cap. In the third year of AHCCCS, six long-term care contractors covered nine counties.

Despite the complexity of relationships among AHCCCS, county governments, long-term care plans, regular plans, and long-term care facilities, AHCCCS regulations and procedures offer little guidance in the area of long-term care. As a result, problems have arisen in coordinating care for long-term care patients and in defining responsibilities and appropriate roles for the many parties involved. In addition, because long-term care is excluded from AHCCCS, the Arizona government receives no Federal funds for long-term care, and the growing cost of these services is a serious financial problem for most counties.

Since the implementation of AHCCCS, there have been ongoing discussions of alternatives to the current system that would involve statewide coverage of long-term care services. Increasing expenditures for long-term care have heightened concern over finding a way to obtain Federal matching funds for these services under Title XIX of the Social Security Act. However, there is still little enthusiasm for implementing a traditional Medicaid program for long-term care. Although discussion is continuing, at the end of the third year of AHCCCS, no changes had been adopted in the Arizona long-term care delivery system.

Although AHCCCS relieved the county governments of sole responsibility for financing indigent health care services, the legislation did not eliminate county obligations altogether. Counties continue to have financial responsibilities associated with indigent health care services in four main areas:

• They fund a portion of the cost of AHCCCS through monthly payments to the program.

• They conduct MI-MN eligibility determinations for AHCCCS.

• They are statutorily obligated by the maintenance-of-effort requirement in the AHCCCS legislation to continue serving people and providing services that would have been covered under their pre-AHCCCS county programs and are not covered by AHCCCS (including long-term care).

• Some provide health care services to other indigent people who are not covered by AHCCCS or by maintenance-of-effort responsibilities (the “notch group”), although not statutorily required to do so.

Although data are not available to measure increases in county expenditures for indigent health care services during the second half of AHCCCS, with few exceptions county governments believe that their expenditures have grown. They relate these increases to growing costs of eligibility determination and the costs of long-term care and other residual care services provided under maintenance-of-effort obligations. Some counties also face increasing costs of caring for the notch group, although a Flinn Foundation report issued in 1985 found that access to health care services for this group had decreased since the implementation of AHCCCS (Flinn Foundation, 1985). The rate of increase in county expenditures may be less than it would have been without AHCCCS, but the implementation of AHCCCS has not provided the financial relief expected by the counties.

County governments that operate hospitals face more difficulties as a result of the implementation of AHCCCS than those that do not because they have encountered more complex problems in adapting their delivery systems to their diminished indigent-care responsibilities. Because a portion of former county patients who are eligible for AHCCCS receive services from private providers rather than the county hospital, some county hospitals have experienced a loss of patients and revenues. As a result, counties have to provide larger subsidies for hospital operations if they do not succeed in cutting expenditures; at the same time, they finance a portion of the cost of caring for former county patients by private providers through their contributions to AHCCCS. Counties that operate hospitals are also likely to serve notch group patients.

During the second 18 months of AHCCCS, four county governments operated hospitals. They have had various degrees of success in adjusting the operation of their hospitals during this time. Three of these counties sponsored AHCCCS plans and successfully bid for AHCCCS contracts. As a result, their hospitals are able to continue serving a portion of their former patient population. Hospitals in two counties, including one that sponsored a plan, faced serious financial difficulties. Although both counties considered divesting themselves of their hospitals, at the end of the second 18 months, both remained in operation.

Procuring providers: Bid process

AHCCCS uses a competitive bid process to select providers for the program and to set prices. By using a competitive bid process, the AHCCCS administration hopes to reduce the cost of caring for the Medicaid population to below the cost of a traditional fee-for-service Medicaid program and stimulate efficiency in delivery of medical care.

There were three competitive bid processes during the first 4 years of AHCCCS. The first bid process covered the first year of the program. The second bid process resulted in contracts that covered the second program year and were renewable for the third year. Virtually all plans with contracts in the second year renewed them for the third year. At the end of the third year, a third bid process resulted in contracts for the fourth program year. These contracts were renewed for the fifth year.

For all three bid processes, program eligibility criteria have remained constant. Service coverage has expanded slightly over time, and a number of program changes have increased the speed with which AHCCCS eligibles became enrolled in a plan. For example, freedom of choice on the part of MI-MN eligibles was restricted by the implementation of automatic assignment to a particular plan using an algorithm developed by the State government.

The bid process has evolved from one that encouraged the formation of new prepaid plans to one that emphasizes the financial responsibility, organizational structure, and provider networks of successful bidders. During the first bid process, 113 bids were received from 50 separate organizations. Partial-service bids were considered, and bidders were required to submit a minimal amount of financial data. In the fourth program year, 46 bids were received from 15 separate organizations. Only full-service bids were considered, and plans were required to submit detailed financial data along with their bids.

Throughout the first three bid processes, plans were asked to submit separate bids by category of eligibility. During the first bid process, the five eligibility categories were AFDC and aged without Medicare, aged with Medicare, blind, disabled, and MI-MN. Recognizing the cost differences between treating the Medicare-eligible and non-Medicare-eligible populations, the bid categories were revised for the second bid process. During the second year they were AFDC, aged and MI-MN with Medicare, aged and MI-MN without Medicare, blind and disabled with Medicare, and blind and disabled without Medicare. For the fourth program year, bid categories were revised again and expanded to seven: AFDC, disabled with Medicare, disabled without Medicare, aged and blind with Medicare, aged and blind without Medicare, MI-MN with Medicare, and MI-MN without Medicare.

In the first and second bid processes, plans were asked to specify both the bid price per eligibility category and the total number of eligibles they were willing to enroll in a plan. Bids were compared across plans using a composite bid rate, which was a weighted average of bids by eligibility category where the weights were the number of eligibles in each category. Statewide weights were used the first year, and county wide weights were used the second year. For the fourth year, plans were no longer asked to specify the number of eligibles they were willing to enroll. Also, although successful bidders still won contracts for all eligibility categories, bid rates were separately evaluated for each category.

During the third program year, a number of plans experienced financial difficulties. Arizona Physicians IPA, Inc. filed for bankruptcy under Chapter 11 of the U.S. Bankruptcy Code; Western Sun Associates, Inc. filed a Chapter 7 bankruptcy; the contract of Health Care Providers of Arizona was terminated by AHCCCS; and the prime contract of El Rio Santa Cruz Neighborhood Health Plan was transferred to Mercy Care.

Arizona Physicians IPA was allowed to continue to provide services to AHCCCS eligibles while it was under reorganization. The plan has settled with its creditors and continues to be a major participant in AHCCCS.

Western Sun's contract was terminated by AHCCCS on June 15, 1985. At that time, Western Sun had 3,633 enrollees in Yuma County and 760 enrollees in La Paz County. A special bid process was issued for Yuma and La Paz Counties, asking for bids covering the period July 15, 1985, to September 30, 1985. Samaritan Health Service/Medical Care Systems won the contract. In August, the contract was sold to Arizona Physicians IPA. Western Sun had originally filed for a Chapter 11 reorganization. However, the plan's biggest creditor, Yuma Regional Medical Center, filed a petition to put the plan in Chapter 7 bankruptcy, reducing the probability that the plan could recover.

AHCCCS terminated its contract with Health Care Providers because of concerns regarding the plan's financial viability and ability to meet its service commitments. Health Care Providers had been a major service provider in Maricopa and Pinal Counties. When AHCCCS terminated its contract, all the enrollees in Pinal County were transferred to other plans. However, in Maricopa County, neither Arizona Physicians IPA nor Maricopa County Health Plan was willing to take on any new MI-MN non-Medicare eligibles without an increase in the capitation rate. Therefore, a special rebid was held in Maricopa County in March 1985 for the MI-MN non-Medicare population.

In August 1985, El Rio, a plan also beset with financial problems, ceased being a contracting AHCCCS plan. It remains operational as a neighborhood health center and continues to participate in AHCCCS as a subcontractor for physicians' and other outpatient services. The El Rio contract, which covered AHCCCS enrollees in Pima County, was transferred to Mercy Care without a new bid process.

Perhaps as a result of the plan financial problems and the program disruptions resulting from them, AHCCCS began to allocate additional resources to monitoring the financial status of the plans. The increased concern for the plans' financial conditions was reflected in the criteria used by AHCCCS to evaluate the fourth-year bids.

The bids for the fourth program year were evaluated using a 200-point scoring scheme. Only 40 points were directly related to the bid prices. The remaining 160 points were related to plan organizational structure (20); provider network (25); operational structure—quality assurance, utilization review, procedures for appointments, referrals, member transition to open enrollment and long-term care, complaints and grievance procedures, member handbook, early and periodic screening, diagnosis, and treatment coverage, and job descriptions (100); and overall proposal quality (15).

It is evident that, as the program has evolved, price has come to play a relatively small role in the determination of the winning bidders. Much more emphasis is given to the organizational structure, financial condition, and provider networks of the plans. This emphasis on nonprice criteria may be in the best interests of the AHCCCS-eligible population in terms of ensuring access to care. However, the shift in emphasis from price evaluation to other performance aspects may result in a longrun decrease in competition. If the request for proposal (RFP) requirements are prohibitively costly for prospective bidders, they may decline to submit bids. The result could be no new entry of plans into the market. Such a reduction in the level of competition may dampen the cost savings potential of AHCCCS and could lead to a more restrictive regulatory environment in the market for health services for the indigent population.

Characteristics of plans

At the beginning of the fourth year of program operation, 15 plans provided services to AHCCCS members. All 15 counties in Arizona had at least one prepaid plan. Three plans of about equal size—Arizona Physicians IPA, ACCESS Patient's Choice, and Maricopa County Health Plan—had almost two-thirds of the enrollees as of September 15, 1985. Arizona Physicians IPA and ACCESS Patient's Choice are statewide independent practice associations that operate in 12 and 9 counties, respectively. Maricopa County Health Plan is a group model health maintenance organization (HMO) sponsored by Maricopa County and serving AHCCCS enrollees in that county. The next largest plans are Mercy Care, a staff model HMO sponsored by a Catholic hospital system, which serves beneficiaries in eight counties, and Pima Health Plan, a group model HMO sponsored by Pima County, which serves AHCCCS beneficiaries in that county.

In Table 1, the distribution of the plans along various dimensions is shown as of September 15, 1985. Plans are categorized as IPA-type, group model, or staff model plans. Of all enrollees, 55 percent are in the seven IPA-type plans, 13 percent in the four staff model plans, and 32 percent in the four group model plans.

Table 1. Arizona Health Care Cost Containment System (AHCCCS) 4th year providers, by percent of enrollees and major features: September 15, 1985.

Plan	Percent of enrollees1	Type	Profit or nonprofit	Payment mode
				Primary care physician	Specialist	Hospital
ACCESS Patient's Choice	21.1	I	P	F	F	D
Arizona Physicians IPA, Inc.	26.5	I	P	C	F	P/A
Comprehensive AHCCCS	1.3	I	N	C	F	C/A
Doctor's Health Plan	0.8	G	P	C	F	C
Dynamic Health Services	1.5	S	P	S/F	F	A
Family Health Plan of Northeastern Arizona	1.2	I	N	FR	FR	C
Gila Medical Services	1.1	S	P	S/F	F	A
Maricopa County Health Plan	20.7	G	N	S	S	P
Mercy Care Plan	7.5	S	N	C	F	D
Northern Arizona Family Health Plan	2.2	I	N	FR	FR	C
Phoenix Health Plan	2.4	I	N	F	F	C/D
Pima Health Plan	7.6	G	N	S	S	D
Pinal General Hospital	2.5	S	N	S/F	F	A
Samaritan Health Service/Medical Care Systems	0.4	I	N	C/F	F	C/A
University Famli-Care	3.2	G	N	S	S	D

¹Total enrollees (125,195) represent people categorically eligible through Federal programs (86,266 Aid to Families with Dependent Children and Supplemental Security Income enrollees) and medically needy or medically indigent eligibles (38,929) served through a prepaid capitated contract with AHCCCS plans. An additional 23,757 AHCCCS eligibles were served through other kinds of AHCCCS contracts. AHCCCS-covered acute medical care services were paid through prepaid capitated contracts or on a capped fee-for-service basis for 3,181 AHCCCS eligibles residing in long-term care facilities. (The 6 long-term care contractors are Maricopa Health Plan, Pima Health Plan, Pinal General Hospital, Dynamic Health Services, Arizona Physicians IPA, and Northern Arizona Family Health Plan.) The Department of Economic Security received a capitated fee, negotiated with AHCCCS, for care of 545 foster children. Plans were reimbursed on a capped fee-for-service basis for 784 Native Americans living on reservations who chose to receive their services from an AHCCCS provider. Finally, AHCCCS administered Federal payments to Indian Health Service facilities for Native American AHCCCS eligibles (19,247) who chose to receive their health care in those facilities. An additional 11,417 people were eligible for AHCCCS but had not yet enrolled in a prepaid plan.

NOTE: Type—I: Independent practice association; G: Group model plan; S: Staff model plan. Profit versus nonprofit—P: Profit; N: Nonprofit. Primary care physician payment mode—C: Capitation payment; F; Fee-for-service; FR: Fee-for-service from a risk pool; S: Salary. Specialist payment mode—F: Fee-for-service; FR: Fee-for-service from a risk pool; S: Salary. Hospital payment mode—A: State government adjusted billed charges (ABC's); C: Capitation; D: Additional discount over ABC's negotiated by the plan for some hospitals; P: Per diem rate.

SOURCE: (McCall et al., 1987).

Enrollees are nearly equally divided between for-profit and nonprofit plans. For-profit plans enroll 51 percent of AHCCCS beneficiaries, and 49 percent are enrolled in nonprofit plans.

With respect to primary care physicians' mode of payment, the majority of the enrollees (63 percent) are in the nine plans that pay their primary care physicians on capitation, fee-for-service, or some combination of capitation and fee-for-service. Three plans (having 32 percent of the enrollees) pay their primary care physicians on salary, and three (serving 5 percent of the enrollees) pay physicians through a combination of salary and fee-for-service.

Specialists are normally paid on a fee-for-service basis. Fee-for-service specialist payment is used by 12 plans (having 68 percent of the enrollees). Only three plans (with 32 percent of the enrollees) pay their specialists on salary.

Four plans pay hospitals with an additional discount over State government adjusted billed charges (ABC's); these four plans have 39 percent of the enrollees. One plan (having 27 percent of the enrollees) pays a combination of per diem rates and ABC's. Another plan (serving 21 percent of the enrollees) pays per diem rates. Three plans (serving 4 percent of the enrollees) pay a capitation rate. Three plans (serving 5 percent of the enrollees) pay ABC's. One plan (serving 2 percent of the enrollees) pays a combination of capitation and additional discount over ABC's. Two plans (serving 2 percent of the enrollees) pay a combination of capitation and ABC's.

In this short description, we show the diversity of organization and governance of the participating AHCCCS plans. Although all plans are capitated by the State government, they have a variety of organizational structures and set up a wide variety of relationships with their participating providers.

Financial status of plans

Financial data were examined for the prepaid health plans that participate in AHCCCS. The primary goals were to assess plan financial performance under AHCCCS and to examine the financial condition of AHCCCS contractors. Evaluation of plan financial performance is important because prepaid health plans are vital components of the AHCCCS delivery system. Therefore, the financial soundness and viability of the plans are critical elements of the AHCCCS demonstration. In addition, a principal cost-containment mechanism of AHCCCS is the use of prepaid capitated financing as the primary method of reimbursing participating plans. Thus, incentives for plan efficiency are particularly important because they affect the overall cost of the program to the State government.

A variety of financial measures and ratios were examined to assess plan financial solvency and profitability. From the evaluation of the solvency measures, it appears that the 15 plans that were awarded contracts for the fourth program year, starting October 1, 1985, are in adequate financial condition. In addition, the plans with the weakest financial results all have financial backers with substantial resources in the event that financial problems develop. However, as discussed in more detail in Chapter VIII of the Second Implementation and Operation Report, several factors preclude drawing definitive conclusions concerning the financial condition of these plans until additional data are available for analysis and evaluation (McCall et al., 1987).

With respect to plan profitability, it appears that the overall profit rate for AHCCCS plans was low during the first 30 months of AHCCCS operations. This condition is a cause for concern, especially if it continues. Most AHCCCS contractors were newly formed prepaid health plans, and it is not unusual for such plans to experience low or negative profit rates as a percentage of revenues during the first 2 or 3 years of operations because of the startup costs associated with a new plan and the development costs for required systems. The key issues are whether these plans can operate profitably after the startup period and whether the capitation rates used for reimbursement are adequate to cover the costs of required services. From the available data for the first 30 months of operations, it appears that there might be cross-subsidization among reimbursement categories. (Most plans make a profit on some categories and suffer a loss on others.) However, further data are required for a complete investigation of this issue.

During the first 3 years of AHCCCS operations, four AHCCCS plans experienced serious financial problems. One plan filed for bankruptcy under Chapter 7 of the U.S. Bankruptcy Code. This plan was liquidated and is no longer in business. Two plans filed for protection under Chapter 11 of the U.S. Bankruptcy Code after incurring large financial losses. One of these plans is inactive, with no revenues or earning assets. The other has since reorganized under new owners and new management. It submitted a bid for the fourth program year and was awarded a contract by AHCCCS for the year starting October 1, 1985. A fourth plan lost its AHCCCS contract as a prime contractor because of its poor financial condition but was allowed to continue participating as a subcontractor.

Plan bankruptcies could have serious negative consequences for AHCCCS as a whole if, as a result, patients did not receive needed services or if other adverse conditions or major disruptions of the system resulted. On the basis of the experience of the third program year, it appears that AHCCCS was able to preserve continuity of patient care and implement an orderly process for the transfer of enrollees from bankrupt plans to other plans. In addition, it appears that AHCCCS was able to take advantage of the lessons learned from the experience with the plans that suffered financial difficulties during the first 3 years of operation. An early warning system was developed to provide program officials with early indicators that a plan is experiencing financial problems so that appropriate corrective actions can be formulated and implemented. However, because this system was not developed until the end of the third program year, its success will depend on experience during the next few years.

Major findings

The second 18 months of AHCCCS were marked by a stabilization of the program's functions and administration. The new leadership that took control of AHCCCS at the end of its first 18 months of operation has provided stable, consistent, and professional management of the program and has made substantial inroads into the main problems seen in the first 18 months.

In its second 18 months of operation, AHCCCS has:

• Initiated meaningful quality assurance reviews and ensured that quality assurance activities are in place at the plan level.

• Collected financial data on the plans to ensure that the program will be able to monitor impending plan problems.

• Reduced eligibility and enrollment time gaps by initiating changes in the enrollment process and developing smoother working relationships with those from whom eligibility data are received.

• Imposed strict eligibility controls and eligibility process review. Doing so has decreased the number of eligible beneficiaries from 190,000 in May 1984 to 160,369 in September 1985. Although this change has decreased the program's financial liability, it may have reduced access to medical care by Arizona's near poor.

AHCCCS has had success in several areas.

• Although four plans experienced serious financial difficulty and two of these were phased out of serving AHCCCS enrollees, the transition of their enrollees to new plans was reasonably smooth.

• Plans that currently have AHCCCS contracts appear, from the financial data available, to be on reasonably sound financial ground. They are becoming increasingly knowledgeable about managing themselves in a prepaid environment, having measures in place to monitor and control utilization and to track expenditures that are incurred but not reported.

• Plans have become more serious about initiating internal cost control mechanisms. Some have initiated prior authorization on all services.

• Many plans continue to be interested in seeking other corporate entities to market to the private sector. To the extent that the medical marketplace is enhanced by these new entrants, the entire climate of health care delivery in the State is enhanced.

In other areas, results have been more mixed.

• An organized system of bidding for the provision of services has been put into place. However, there were criticisms that the system, although highly structured, was not adequately described to the potential bidders. Also, the points assigned in the evaluation process gave a heavy weight to bidders' qualifications and a small weight to price. Given the high cost of preparing a bid, the competitive nature of the marketplace might be substantially threatened if such a system continues.

• The AHCCCS administration has begun a major effort to redesign its management information system to accommodate the special needs of a program that administers prepaid plans. This is a laudable effort given the increased emphasis by public programs on reimbursing their providers under capitated arrangements and the lack of any working systems. However, the current AHCCCS MIS is not used as effectively as it might be, and efforts to collect encounter data from the plans have not been given as much attention as have other issues during the second 18 months of the program. As a result, the MIS remains AHCCCS's most critical implementation problem.

• The division of responsibility between acute and long-term care also remains an area in which little has been accomplished in the second 18 months of the program. Although there are significant problems in attempting to capitate long-term care services, other approaches to coverage under a unified program management could result in more effective health care delivery than is currently received by the AHCCCS population.