Health Policy for Low-Income People in Arizona

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

---

About the Series

Assessing the New Federalism is a multi-year Urban Institute project designed to analyze the devolution of responsibility from the federal government to the states for health care, income security, employment and training programs, and social services. Researchers monitor program changes and fiscal developments. In collaboration with Child Trends, Inc., the project studies changes in family well-being. The project aims to provide timely, nonpartisan information to inform public debate and to help state and local decisionmakers carry out their new responsibilities more effectively.

Key components of the project include a household survey, studies of policies in 13 states, and a database with information on all states and the District of Columbia, available at the Urban Institute's Web site. This paper is one in a series of occasional papers analyzing the information from these and other sources.

---

Contents

Highlights of the Report
Thumbnail Sketch of Arizona
The Arizona Health Care Cost Containment System: An Overview

Before AHCCCS: Holding on to a County-Based System
Enactment of AHCCCS: A Unique Federal-State Partnership
Emerging as a Managed Care Model: AHCCCS in the 1990s
The State's Unsuccessful Effort to Expand Eligibility
The AHCCCS Program: Some Basic Information

The AHCCCS System of Managed Care

Marketing and Enrollment
Paying Managed Care Plans: The Competitive Bidding Process
Quality of Care

Arizona Child Health Insurance Program
Arizona's Efforts to Increase the Availability and Affordability of Private Insurance

Health Care Group of Arizona
Arizona Premium Sharing Program

The Medical Safety Net in Arizona: Coping with the Rising Numbers of Uninsured
The Arizona Long-Term Care System

Who Is Eligible to Receive ALTCS Benefits?
Which Health Plans Participate in ALTCS?
The ALTCS Delivery System
The Nursing Home Industry in Arizona
Home- and Community-Based Services in Arizona
Paying the ALTCS Bill: Controversy over the State Share
Links between ALTCS and Medicare
Evaluating the ALTCS System: Literature on Cost and Quality

The Arizona Behavioral Health System
Conclusion Notes About the Author

---

Highlights of the Report

Arizona is well known for its innovative Medicaid program, called the Arizona Health Care Cost Containment System (AHCCCS). The program's national reputation is rooted in three features. First, the relationship between AHCCCS and the federal government is unique. AHCCCS began in 1982 as a Medicaid demonstration pilot, exempt from many of the federal rules that govern other state programs. Seventeen years later, it continues to operate under this special status. Second, most AHCCCS beneficiaries have always been required to enroll in managed care. AHCCCS operates the nation's oldest and most comprehensive Medicaid managed care initiative. Third, the system under which managed care plans compete for Medicaid contracts is considered a national model. Observers often attribute the state's ability to contain Medicaid costs to the implementation of its system of competitive bidding.

Despite its innovative Medicaid program, Arizona is struggling with a major health care crisis. Over the past seven years, the percentage of nonelderly state residents without health insurance has grown from roughly 21 percent to just over 28 percent. Arizona now ranks first in the nation in the percentage of persons without health insurance. The state has large numbers of Hispanics (24.7 percent of the state population) and Native Americans (5.6 percent of the population), both groups with higher-than-average rates of poverty, lack of health insurance, and poor health status.

There are several explanations for the state's growing health care crisis. The most important factor is the state's economy. Although unemployment is quite low (less than 4 percent), most recent job growth is in the service sector, especially in the tourist industry. These jobs are relatively low paying and often do not provide health insurance. Indeed, only two states (Arkansas and New Mexico) have a lower percentage of persons with employer-based health insurance coverage. At the same time, Arizona's Medicaid program, while innovative, is not particularly generous, especially for adults with income below the federal poverty level (FPL). Ironically, federal officials have rejected a state request to amend its waiver to liberalize Medicaid eligibility on the ground that the expansion would not be budget neutral.

Welfare reform is making the situation even more problematic. Large numbers of former welfare beneficiaries are still eligible for Medicaid but have not enrolled, either because they are unaware of their continuing eligibility or because they are deterred by the administrative burden of applying. Indeed, even AHCCCS managers concede that the state needs to dramatically expand its outreach and education efforts. Some efforts to do so are now under way.

Arizona officials have enacted several programs in response to the health insurance crisis. The first is a state-administered child health insurance initiative, administered under the umbrella of the national Child Health Insurance Program. This program, known as KidsCare, covers children in families with income below 150 percent of the FPL (this level will rise to 200 percent in October 1999). So far, however, enrollment is well below expectations. Here again, a big problem is lack of outreach and education, although state officials insist that recent marketing efforts will minimize this problem.

Two initiatives target working adults: The Premium Sharing Program provides subsidized insurance for persons with income below 200 percent of the FPL, and the Health Care Group provides low-cost insurance to small businesses around the state. However, these programs, too, are struggling with low enrollment and disappointing outcomes.

Finally, the state has distributed funds generated by a tobacco tax to safety net health care providers to subsidize care provided to the uninsured. These grants, along with federal subsidies for community health centers and for hospitals that serve a disproportionate number of the poor, have enabled the safety net to continue its mission. At the same time, many of these providers have adopted managerial efficiencies that have cut costs. Nonetheless, the pressures on the safety net remain enormous. The federal government is reducing its subsidies to safety net providers. Several of the managed care plans in AHCCCS are struggling to make a profit on their managed care operations and are looking to reduce expenditures. The rising number of uninsured persons means a constant increase in the level of uncompensated care, with no assurance that tobacco tax revenue will keep pace.

The state's behavioral health system also seems to be in crisis. The main problem is unacceptable conditions in the large state mental hospital. Conditions are so poor that federal officials recently took away the institution's accreditation. Shortly thereafter, Governor Jane Hull announced that her number one health care priority is to address the problems in this institution. Governor Hull hopes to persuade the state legislature to allocate funds from the tobacco settlement to build a new facility.

The state's long-term care system seems to be in stable condition. The linchpin of the system is the Arizona Long-Term Care System (ALTCS), which is the long-term care component of the AHCCCS program. Beneficiaries receive all Medicaid-covered services from an ALTCS managed care contractor. There is one such contractor in each of the state's 15 counties (unlike the acute care program, which features several health plans and a vigorous system of competitive bidding). The contractor develops a network of providers that includes primary care physicians, hospitals, home health agencies, nursing homes, and others. Case managers who work for the contractor then help the beneficiary and his or her family choose the appropriate level of care. Because the contractor receives a capitation payment from the state, it has an incentive to place large numbers of clients in the (less expensive) home- and community-based system. As a condition of its federal waiver, however, no more than 50 percent of ALTCS beneficiaries may reside in home- and community-based settings.

Four issues dominate the ALTCS agenda. First, ALTCS beneficiaries must be sufficiently disabled to be "at risk" of receiving long-term institutional care. By all accounts, ALTCS officials interpret this requirement more restrictively than do other states. One result is that ALTCS beneficiaries are more disabled than their counterparts in other states. This restrictive eligibility criterion has led to complaints that needy persons are wrongfully denied long-term care benefits. State officials insist that nearly all of the truly needy are covered and that the restrictive criterion enables the state to concentrate scarce resources on those most in need.

Second, state officials are anxious to infuse health plan competition into the ALTCS system. For this reason, effective in 2001, officials will eliminate the requirement that county governments in Maricopa (i.e., Phoenix) and Pima (i.e., Tucson) administer the plan in their counties. State officials also will eliminate the requirement that three other counties have the right of first refusal for the administration of the plan in their region. Under the new rule, there will be open competition for the ALTCS contract in all of the state's 15 counties.

Third, before 1998, the state refused to share in the cost of ALTCS: The counties were required to pay the nonfederal share. In late 1997, however, the legislature revised the long-term care payment formula to require the state to pay 50 percent of the nonfederal share of any amount above what the counties paid in 1997/1998. Despite the change, the intergovernmental division of payments remains controversial.

Finally, the state has struggled for years to encourage dual eligibles—persons who receive Medicaid and Medicare—to receive all of their medical services from a single managed care contractor. In the mid-1990s, the state requested federal permission to require most beneficiaries to receive all Medicare-covered services from their ALTCS contractor. Federal officials denied the request on the grounds that Medicare beneficiaries cannot be required to join managed care organizations. Since that time, AHCCCS officials have decided to pay the Medicare cost-sharing (copayments and deductibles) for dual eligibles only when clients receive care from an AHCCCS managed care plan. This requirement, and the more general issues surrounding dual eligibles, remain extremely controversial.

Thumbnail Sketch of Arizona

Over the past decade, Arizona has experienced a surge in population. Between 1990 and 1998, the number of residents grew to 4.6 million, a 21 percent increase,¹ making the state the 21st most populous in the nation.² Much of the state's population growth is concentrated in Maricopa County (Phoenix and its suburbs); this region recently passed Las Vegas as the fastest-growing community in the country. The state's warm climate and low unemployment rate³ are a magnet for the young. At the same time, Arizona ranks second (behind Florida) in the number of elderly persons migrating from other states.⁴

As Arizona's population has grown, so, too, has the number of state residents without health insurance. The percentage of uninsured nonelderly residents has increased from 20.8 percent in 1989-90 to 28.1 percent in 1996-97.⁵ Arizona now has the dubious distinction of having the highest percentage of uninsured persons in the nation. As shown in table 1, the state also struggles with high rates of poverty and a lower-than-average per capita income. In 1996, for example, more than 18 percent of the state's population lived in poverty, compared with less than 14 percent nationwide.⁶ In 1997, the state's per capita income of $22,364 ranked 36th in the nation, well below the national average of $25,598.⁷ Finally, the state has large numbers of Hispanics (24.7 percent of the state population)⁸ and Native Americans (5.6 percent of the population),⁹ both groups with higher-than-average rates of poverty, lack of health insurance, and poor health status.

These trends (low unemployment, low wages, and high numbers of uninsured) are consistent with the rapid growth in the service sector of the state's economy. The ongoing boom in the state's hotel and tourism industry generates jobs for large numbers of new state residents. The boom also aids the state's welfare reform initiative. The availability of entry-level service-sector jobs enabled state officials to reduce the number of welfare beneficiaries by more than 60 percent between 1996 and 1998. By and large, however, these service-sector jobs are low paying and do not include health insurance. Indeed, Arizona ranks 48th in the nation in the percentage of persons with employer-based health insurance coverage (60 percent, ahead of only Arkansas and New Mexico).¹⁰ Nor have most former welfare beneficiaries enrolled in the state's transitional Medicaid initiative; instead, most of the former beneficiaries have joined the ranks of the uninsured.

Table 1  State Characteristics
	Arizona	United States
Sociodemographic
Population (1998)a (in thousands)	4,575	270,002
Percent under 18 (1994-95)b	28.2%	26.8%
Percent 65+ (1994-95)b	12.6%	12.1%
Percent Hispanic (1994-95)b	24.7%	10.7%
Percent Non-Hispanic Black (1994-95)b	3.3%	12.5%
Percent Non-Hispanic White (1994-95)b	68.5%	72.6%
Percent Non-Hispanic Other (1994-95)b	3.5%	4.2%
Percent Noncitizen Immigrant(1994-95)b	3.5%	4.2%
Percent Nonmetropolitan (1994-95)b	18.1%	21.8%
Population Growth (1995-96)c	2.9%	0.9%
Economic
Per Capita Income (1997)a	$22,364	$25,598
Percent Change in Per Capita Personal Income(1995-96)d	5.1%	4.6%
Percent Change in Personal Income (1995-96)d	8.1%	5.6%
Employment Rate (1997)e,f	60.8%	63.8%
Unemployment Rate (1997)f	3.7%	4.5%
Percent below Poverty (1995-96)a	18.3%	13.8%
Percent Children below Poverty (1994)g	24.9%	21.7%
Health
Vaccination Coverage of Children Ages 19-35 Months (1997)a	74.0%	78.0%
Low-Birth-Weight Births (<2,500 g) (1995)h	6.8%	7.3%
Infant Mortality Rate (Deaths per 1,000) (1995)i	7.6	7.2
Premature Death Rate (Years Lost per 1,000) (1995)j	51.4	46.7
Violent Crimes per 100,000 (1996)k	631.5	634.1
AIDS Cases Reported per 100,000 (1997)a	9.8	22.3
Political
Governor's Affiliation (1998)i	R
Party Control of Senate (Upper) (1997)m	12D-18R	22D-38R
a. JoAnn Lamphere, Normandy Brangan, Sharon Bee, and Rafael Semansky. Reforming the Health Care System: State Profiles 1998 (American Association of Retired Persons 1998). b. Two-year concatenated March Current Population Survey (CPS) files, 1995 and 1996. These files are edited by the Urban Institute's TRIM2 microsimulation model. Excludes those in families with active military members. c. U.S. Bureau of the Census. Statistical Abstract of the United States: 1997 (117th edition). Washington, D.C., 1997. 1995 population data as of July 1. d. Bureau of Economic Analysis, U.S. Department of Commerce, January 1998. e. U.S. Department of Labor. State and Regional Unemployment, 1997 Annual Averages. USDL 98-78. Washington, D.C., February 27, 1998. f. Employment rate is calculated using the civilian noninstitutional population age 16 years and over. g. CPS three-year average (March 1994-March 1996, where 1994 is the center year) edited using the Urban Institute's TRIM2 microsimulation model. h. S.J. Ventura, J.A. Martin, S.C. Curtin, and T.J. Mathews. "Advance Report of Final Natality Statistics, 1995." Monthly Vital Statistics Report 45 (11,supp). Hyattsville, MD.: Public Heath Service, 1997. i. National Center for Health Statistics. " Births, Marriages, Divorces and Deaths for June 1996." Monthly Vital Statistics Report 45 (12). Hyattsville, MD.: Public Heath Service, 1997. j. Rate was calculated using years of potential life lost from age 65 (National Center for Health Statistics. Multiple Cause of Death Mortality Tapes, 1995) as the numerator and population estimates (U.S. Bureau of the Census. ST 96-1. Estimated of the Population of States: Annual Time Series, July 1, 1990 to July 1, 1996) as the denominator. k. U.S. Department of Justice, FBI. Crime in the United States, 1996. September 28, 1997. l. National Governor's Association. Governors, Political Affiliations, and Terms of Office, 1998. January 15, 1997. M. National Conference of State Legislatures. D indicated Democrat and R indicated Republican.

State policymakers recently have enacted several efforts to reduce the number of uninsured. The KidsCare program is Arizona's response to the national Child Health Insurance Program (CHIP); the Premium Sharing Program is a state-funded effort to subsidize health insurance for persons with incomes below 200 percent of the federal poverty level (FPL) (400 percent for the chronically ill); and the Arizona Health Care Group offers a state-administered high-risk pool for those in the small-business community. Generally speaking, however, these initiatives enroll relatively few of the uninsured. The state's primary vehicle for providing health care to the poor remains the state's unique Medicaid program, the Arizona Health Care Cost Containment System (AHCCCS).

The Arizona Health Care Cost Containment System: An Overview

Arizona occupies a unique place in the history of the nation's Medicaid program. The story begins with the state's longtime refusal to enact a Medicaid initiative. Arizona was the last state in the nation to join the Medicaid bandwagon and did not do so until 1982. For the next decade or so, Arizona's Medicaid program was considered unique and controversial. Long before Medicaid managed care was fashionable, Arizona required nearly all beneficiaries to enroll in managed care. Long before many policymakers had even heard of Medicaid demonstration waivers, Arizona's entire Medicaid program was run as a demonstration initiative. Over time, however, Arizona has shifted from outlier to trendsetter. The state is no longer considered regressive or revolutionary; other states now seek to emulate much of the Arizona model.

Before AHCCCS: Holding on to a County-Based System

Before the congressional enactment of Medicaid in 1965, the federal government and the states made only minimal efforts to ensure that the poor had access to health care. The assumption was that local governments would provide a medical safety net to the so-called deserving poor.¹¹ In response to this mandate, local governments in Arizona and elsewhere established public hospitals, public health clinics, and a range of other programs. Despite these efforts, however, the health care system available to the nation's poor was generally inadequate. Without federal and state funding (and leadership), the medical safety net was exceedingly thin.

The enactment of Medicaid was intended to improve the health care system for the poor. The model was straightforward: State governments would establish health insurance programs for low-income populations, and, as long as the programs met minimal federal requirements, the federal government would fund between 50 percent and 80 percent of the cost. In most of the nation, the lure of federal dollars proved irresistible, and by the early 1970s every state except Arizona had implemented a Medicaid initiative. In Arizona, however, state officials argued that the benefit of federal dollars was outweighed by the burden of federal Medicaid requirements and the cost to the state treasury. The state has a strong antigovernment political culture, and Medicaid was perceived by many to be another example of big government run amuck.

State officials decided instead to maintain the county-operated and -funded system of indigent care. While county officials were generally unhappy with the decision, the level of discontent was muted until 1980, when the state legislature imposed stringent limits on the counties" ability to raise property taxes. This restriction, along with rising health care costs, created fiscal turmoil in several counties. One county, Santa Cruz, nearly went bankrupt; for months it gave out warrants (IOUs) instead of paying bills. In this environment, county officials began a vigorous campaign for a state Medicaid program to relieve their fiscal problems.

Enactment of AHCCCS: A Unique Federal-State Partnership

As the pressure grew for an Arizona Medicaid program, so too did the state's resolve to maintain control over the structure of any Medicaid initiative. The extent of these differences became clear in the state's 1981 legislative session. During that session, the legislature enacted the AHCCCS and designated it as the state's Medicaid program. The legislators knew, however, that AHCCCS differed in three important ways from what was permissible under a traditional Medicaid program. First, AHCCCS required nearly all beneficiaries to enroll in a competitively bid managed care system—most clients would not have the freedom to choose any willing provider. Second, AHCCCS did not cover long-term care services. Third, AHCCCS did not cover behavioral health services.

In early 1982, Arizona officials requested federal permission, under Section 1115d of the Social Security Act, to operate AHCCCS as an approved Medicaid demonstration pilot. In July 1982, the Reagan administration, anxious to demonstrate its commitment to state flexibility, agreed to grant the request, thereby waiving a host of otherwise applicable federal requirements. The state hired a private contractor (the MCAUTO Systems Group) to administer the program. In October 1982, AHCCCS began to provide health insurance for low-income state residents.

Emerging as a Managed Care Model: AHCCCS in the 1990s

By all accounts, there were significant problems with the initial implementation of AHCCCS. The MCAUTO Systems Group was neither experienced nor adept at handling Medicaid matters, and there were significant administrative gaps, ranging from clients unable to obtain clear answers about eligibility to health plans not paid in a timely fashion. There also was concern about the fiscal and organizational capacity of some participating health plans. Several plans were formed virtually overnight and had inadequate provider networks and insufficient financial reserves. Finally, there was beneficiary resentment over the managed care requirement: Managed care in the early 1980s was not mainstream care.

By early 1984, AHCCCS was attracting national attention. Federal regulators were threatening to revoke the state's 1115 waiver. State officials responded by firing the MCAUTO Systems Group. Rather than hire another contractor, however, the state created a new state agency and vested it with broad bureaucratic autonomy. For many years, for example, the AHCCCS agency was able to hire staff without going through the regular state personnel requirements.12 Within weeks, the new state agency terminated the contracts of two health plans and imposed new reserve requirements on others. Within months, the initial implementation problems were over. Since that time, AHCCCS officials have spent more than the average state on program administration and have generally received high marks for their efforts.

Over time, the AHCCCS system of managed care has become a national model. Even as other states experimented with managed care, Arizona always remained one step ahead of the curve. In 1989, for example, when the state added long-term care to its benefit package, it also required that long-term care services be provided through managed care organizations. Arizona is the only state in the nation with such a requirement. Similarly, when the state phased in behavioral health in the early 1990s, it required that behavioral health services be provided through managed care.

The state's Unsuccessful Effort to Expand Eligibility

When Arizona established its Medicaid program, it also established a small state-funded insurance program for adults with income just above the Medicaid eligibility levels. The Medicaid eligibility level for single, nondisabled adults is 32 percent of the FPL. The state-funded initiative covers adults with income below 40 percent of the FPL. Both programs are considered part of the AHCCCS initiative; the only difference is that the federal government pays 65.6 percent of the cost of caring for Medicaid beneficiaries but does not contribute at all to the cost of caring for other AHCCCS enrollees.

In March 1995, AHCCCS officials requested federal permission to increase the Medicaid eligibility criteria for single adults to 100 percent of the FPL. The proposal would move Arizona's eligibility criteria from one of the least generous in the nation (at 32 percent of the FPL)¹³ to one of the most generous. The motivation was twofold: first, to convert enrollees in the state-funded program to the Medicaid program, and, second, to provide health insurance for large numbers of uninsured persons. In an effort to avoid unexpected costs, AHCCCS officials also sought to cap enrollment in the expansion initiative to 150,000 persons.

Before federal officials could respond to the waiver expansion request, the Arizona legislature itself quashed the proposal. Legislators were not convinced that the enrollment cap would contain the cost of the initiative, and they were dismayed that AHCCCS officials had submitted the proposal without seeking legislative guidance. For these reasons, the legislature ordered the agency officials to withdraw the expansion proposal.

Supporters of the expansion initiative, including the state's medical association, gathered enough signatures to put the proposal to the voters in a referendum. The goal was to use the referendum process to override the legislative opposition. The strategy worked. In November 1996, the state's voters approved Proposition 203 by a 71 percent to 29 percent vote, authorizing the expansion, as long as federal officials approved and contributed to the cost. Early the next year, AHCCCS officials resubmitted their waiver expansion request to the Health Care Financing Administration (HCFA).

Perhaps surprisingly, HCFA officials rejected the resubmitted application. Federal regulators raised two concerns. The first problem was that the request was not budget neutral. State officials sought to circumvent this requirement by claiming a credit for the years of savings that the AHCCCS system has already generated: Since its inception, AHCCCS has arguably saved the federal treasury more than $600 million by having a system of managed care instead of fee-for-service medicine. HCFA officials denied the request for retroactive credit. The second problem was the state's effort to impose a cap on the number of expansion slots: Federal officials are reluctant to allow states to cap the number of Medicaid beneficiaries. Despite the rejection, AHCCCS officials did not abandon the waiver application. State officials have instead reworked the numbers and resubmitted the request. As of mid-1999, the negotiations are ongoing.

The AHCCCS Program: Some Basic Information

AHCCCS enrollment grew rapidly from the mid-1980s until the mid-1990s, increasing from 144,450 enrollees in July 1985 to 476,117 in July 1996. The increase was primarily due to eligibility expansions for children and pregnant women.¹⁴ Since then, however, there has been a steady decline in program enrollment. By July 1998, the number of beneficiaries had fallen to 431,047. The decrease is primarily due to reductions in welfare caseloads. In July 1996, 206,959 welfare beneficiaries were on the Medicaid rolls; two years later that number had dropped to 145,156. Moreover, most former welfare recipients do not sign up for the transitional Medicaid program; rather remarkably, there are fewer enrollees in the transitional program in 1999 than there were in 1996.¹⁵

Table 2 shows that as the number of Medicaid beneficiaries has declined, so has the growth of the state's Medicaid expenditures. Between 1992 and 1995, for example, AHCCCS expenditures grew at an average of 12 percent annually. Between 1995 and 1997, however, expenditures grew by only 4 percent. This trend is consistent with Medicaid programs in the rest of the nation: Across the United States, Medicaid spending increased 9.7 percent annually between 1992 and 1995 but only 3.2 percent between 1995 and 1997.

Although Arizona's overall spending patterns track national trends, the state differs from others in important respects. First, Arizona spends far less per enrollee than the national average ($2,384 versus $3,581 in 1997). However, comparisons across all types of enrollees can be misleading. As shown in table 3, Arizona spends about the same as the national average on the elderly. Spending on adults and children is 12 percent and 6 percent, respectively, below the national average, but it is the low level of spending on the blind and disabled that contributes most to the state's lower overall expenditures. In 1997, Arizona spent $6,640 per blind or disabled enrollee, 25 percent below the average of $8,841 for the United States. Why this occurs is not clear.¹⁶ Second, Arizona's Medicaid expenditures as a percentage of overall state spending are well below the national average (12.4 percent versus 20 percent).¹⁷ Third, Arizona spends more on administration than does the average state. (Interestingly, however, willingness to spend on administration is cited by many as a key factor in the state's low overall costs).¹⁸

Table 3  Medicaid Expenditures per Enrollee by Eligibility Group, Arizona and United States   Arizona, 1997 United States, 1997 Total $2,384 $3,581 By Group Elderly $10,609 $10,804    Cash 4,611 5,965    Noncash 15,499 14,615 Blind & Disabled $6,640 $8,841    Cash 5,897 7,744    Noncash 10,214 12,759 Adults $1,670 $1,879 Children $1,086 $1,156 Source: The Urban Institute, 1999. Based on HCFA 2082 and HCFA 64 data.

The AHCCCS System of Managed Care

AHCCCS has developed three separate managed care systems. The first and largest is made up of 11 health plans that provide primary and acute care services to beneficiaries who do not receive long-term care services. The second system is a group of five behavioral health firms that provide mental health services to this same population.¹⁹ The third system comprises eight health plans that care for just over 25,000 recipients of long-term care services. These plans provide clients with all services: case management, primary and acute care, behavioral care, and long-term care.

The large majority of AHCCCS beneficiaries are enrolled in one or more of these managed care systems. In late 1997, for example, just over 81 percent of all AHCCCS enrollees were in managed care. The only groups that are permitted to stay in fee-for-service medicine are Native Americans, children in foster care, persons residing in institutions other than nursing homes, and persons eligible to receive only emergency services.

Marketing and Enrollment

Before 1998, new AHCCCS beneficiaries received a managed care enrollment packet in the mail and had to select a plan within 16 days. During this period, beneficiaries were covered by the state's small fee-for-service Medicaid program. Beneficiaries who did not select a plan in a timely fashion were assigned to one. Between 60 percent and 70 percent of beneficiaries chose a plan and the rest were assigned. In making these assignments, state officials were guided by four criteria: (1) family continuity (other members enrolled with a particular provider), (2) ZIP code (to encourage geographic access), (3) quality (based on the quality score the plan received during the contracting process), and (4) price.

In 1998, however, the state changed the marketing and enrollment system. The goal was to reduce the number of persons in fee-for-service. Under the new system, AHCCCS encourages Medicaid applicants to select a health plan before the completion of the eligibility application process. Persons who do not make a selection and who are found to be program eligible are assigned to a plan without the 16-day grace period. Most beneficiaries do not voluntarily select a plan; the Medicaid autoassignment rate has grown to 70 percent. AHCCCS also assigns all state-funded beneficiaries to health plans; these clients do not have any freedom of choice. As a result, the autoassignment rate for the entire AHCCCS population (Medicaid and state-funded) is close to 90 percent.

The rise in the autoassignment rate has raised concern among consumer advocates and some state officials. The concern is accentuated because the state has a 12-month lock-in period: Clients who do not opt out of their plan within the first 16 days are generally required to stay in the plan for 12 months.²⁰

The state also recently changed the process by which beneficiaries renew their managed care membership. Until this year, the state had an open enrollment period every August. Clients received a mailing in July and had 16 days to change their plan. Under the new system, clients still can change plans annually, but this selection takes place on the anniversary of the date they joined AHCCCS, not during an annual open enrollment period. The state changed the system to spread the flow of new clients to health plans throughout the year and to lower AHCCCS administrative costs.

While the enrollment process has changed, the marketing process has not. Health plans are not allowed to engage in any direct marketing. Instead, the state's centralized mailing service sends each client a packet that contains informational brochures from each health plan, along with a letter from the state explaining the various options. Clients with questions about their managed care options can visit a Medicaid office and speak to an eligibility worker, or they can call a toll-free telephone line. By most accounts, however, relatively few clients take advantage of these options.

Paying Managed Care Plans: The Competitive Bidding Process

Arizona uses a system of competitive bidding to select health plans for the AHCCCS market. The state scores the bids on four factors: (1) provider network (35 percent of the overall score), (2) capitation rate (30 percent), (3) programs and services (20 percent), and (4) organizational structure (15 percent). State officials will work with the plans to increase their overall score (perhaps by proposing a lower capitation rate), but little real negotiation occurs.²¹ Successful bidders receive five-year contracts (before 1997, the contracts were for two or three years). After the first year, the state increases the capitation rate to account for inflation and any programmatic or legislative changes.

Before 1992, there was little competition for AHCCCS contracts. The largest plans were county sponsored (in Maricopa and Pima) and the only other plans were provider sponsored. This pattern began to change in 1992, when Inter-Group became the first commercial health maintenance organization (HMO) to receive an AHCCCS contract. By 1994, interest in the AHCCCS market reached an all-time high: AHCCCS plans that year reported overall profits of $56 million (or 6.5 percent of gross revenue), and the rest of the industry took notice.

In 1994, for the first time, there was significant competition for AHCCCS contracts: 21 health plans submitted bids, but only 14 received the three-year contracts. Among the successful bidders were four commercial HMOs: Blue Cross, Cigna, Inter-Group, and Regional. Among the unsuccessful bidders were two nonprofit plans that had previously had AHCCCS contracts. According to state officials, the successful plans generally submitted bids with low capitation rates; the HMOs were anxious to enter the market and were willing to undercut the competition to do so. At the same time, state officials were anxious to encourage commercial HMO participation, and they were pleased that the average capitation rate fell by approximately 11 percent.

Between 1994 and 1997, however, many of the participating HMOs began to rethink their commitment to the program. The main complaint was that the rates (which the plans themselves had bid) were too low, especially in rural communities. Medicaid managed care was more costly and more difficult than some of the new entrants had predicted. This problem was exacerbated by the changing nature of the Medicaid population: There were fewer clients (making it harder to spread risk) and the average client was more costly, since many of those cut from the rolls were healthy children. As a result, several plans lost a lot of money in their first foray into the Medicaid market. Blue Cross, for example, lost $30 million on its Medicaid contract and threatened to exit the market unless rates were significantly increased.

Table 4  Managed Care Enrollment by Health Plan, April 1, 1999 Plan Number of Enrollees Arizona Physicians IPA 114,946 Mercy 63,844 Phoenix Health Plan 42,340 Health Choice 32,147 Maricopa 22,911 Cigna 21,810 University Physicans 10,859 Arizona Health Concepts 8,094 Pima 5,788 Family Health Plan 2,574 Doctors Health Plan 1,704 Source: Materials received from AHCCCS.

It is not easy to evaluate the adequacy of Medicaid capitation rates. A recent study suggests, however, that Arizona is not a low payer in comparison with other states. The study compared capitation rates in 36 states: Arizona ranked 12th with an average adjusted rate of $141.46, which is significantly higher than the national average of $129.32.²² Moreover, the regional variation in the rates (between urban and rural areas) is relatively small, casting some doubt on the validity of complaints about AHCCCS rates in rural areas.²³

National comparisons aside, however, the Arizona HMO industry is not convinced it can make money in the Medicaid market. At the same time, many of the new entrants are also losing money in the commercial and Medicare markets. These trends persuaded some plan officials of the need to concentrate on the core of their business—the commercial market—and to leave the Medicaid market. Both Inter-Group and Regional Health Plan cited losses in these other markets as the basis for their decision to exit the Medicaid program.

The rate controversy was the key issue during the 1997 bidding process. State officials, worried that there would not be enough bids in rural counties, reduced the number of service areas from 15 to 9. As predicted, there were fewer bids, especially in rural communities. Statewide, the number of bidders dropped from 21 (in 1994) to 14. Some plans, including Cigna and Arizona Physicians Independent Practice Association (IPA), rejected the state's final offer in some rural communities and accepted contracts elsewhere. Other plans, such as Blue Cross, exited the market altogether.24 In some regions, such as Tucson, there still is significant competition (only five out of nine bidders received contracts). Nonetheless, the trend is toward fewer plans and less competition. Table 4 lists the health plans now in the market and their total enrollment as of April 1, 1999.

Quality of Care

Across the nation, Medicaid officials struggle to determine the impact of managed care on the quality of care received by Medicaid beneficiaries. The task is especially difficult in Arizona, which has never had a large, traditional fee-for-service program with which to compare the managed care initiative. In assessing quality, state officials rely instead on four main pieces of evidence. First are a series of studies by Nelda McCall and colleagues that examined the quality of care in AHCCCS and concluded that the quality was generally good.²⁵ Second is a study performed by the U.S. General Accounting Office (GAO) that reached a similar conclusion.²⁶ Third are satisfaction surveys, conducted by both AHCCCS and the Flinn Foundation, which suggest that more than three-quarters of the respondents think their health plan provides them with good or very good care.²⁷ Fourth are data that demonstrate that state residents (including those on AHCCCS) have fewer hospital admissions, emergency room visits, and hospital outpatient visits than is the national norm.²⁸

The McCall studies are especially important to the debate. McCall and her colleagues compared the care provided to AHCCCS beneficiaries with that provided to Medicaid beneficiaries in New Mexico. The authors concluded that AHCCCS beneficiaries received better well-child and primary care than their fee-for-service counterparts but were less likely to use institutional and specialty care services.²⁹ The GAO reached a similar conclusion, noting that the state's emphasis on cost containment "appears not to have adversely affected the care provided to Arizona Medicaid beneficiaries."³⁰

Nonetheless, the evidence on quality is hardly conclusive. The positive conclusions are based primarily on a comparison with one state (New Mexico) more than a decade ago. Moreover, even state officials acknowledge ongoing problems with aspects of the program. By all accounts, for example, the percentage of AHCCCS beneficiaries who receive adequate prenatal care is well below the national norm. This problem prompted state officials to enact the Baby Arizona program to encourage pregnant beneficiaries to obtain needed care. Moreover, there is significant beneficiary dissatisfaction with the dental care program. State officials have responded by creating a dental care task force to consider ways to increase the number of participating dentists and dental utilization.³¹

Despite these problems, however, most consumer advocates seem to agree that the AHCCCS beneficiaries generally receive high-quality medical care. The key factor is that more than 80 percent of the state's physicians accept AHCCCS patients. At the same time, AHCCCS officials receive high marks for their oversight of the quality of care provided by participating health plans. Indeed, in a recent study of managed care in five states, Arizona had by far the most sophisticated system of collecting encounter data and distributing comparative information on plan performance.³² The high rate of physician participation guarantees good access to care and makes it likely that client satisfaction will remain high.

Arizona Child Health Insurance Program

The Children's Health Insurance Program (CHIP), enacted by Congress as part of the Balanced Budget Act of 1997, provides states with $20.3 billion over five years³³ to expand insurance programs for children under the age of 19. Arizona's allotment of the CHIP budget is $114 million annually. To receive this full allotment, however, the state needs to contribute $36 million in state general revenues. Under this division of payments, the federal government will pay 75 percent of the Arizona CHIP (substantially more than its 65.6 percent share of the state's Medicaid bill).

Following the enactment of CHIP, Arizona lawmakers engaged in a sharp debate over whether the state should participate in the new federal initiative. Supporters noted that the state had nearly 300,000 uninsured children and had the fourth- highest percentage of uninsured children in the nation. Supporters also noted that the federal government would pay 75 percent of the bill and that the state could use tobacco tax dollars to fund the state share. Opponents focused on the federal strings that accompany any federal dollars and cited the state's maverick heritage. Other issues included the implications of expanding the government-run managed care system (AHCCCS) and the possibility that government funds would be used in school-based health clinics to support family planning services.

The battle over child health led to an effort by conservative Republicans to defeat Governor Jane Hull in her effort to be reelected because of her alliance with legislative Democrats in support of the child health initiative.³⁴ Eventually, however, Governor Hull defeated the insurgency within her own political party and persuaded the legislature to enact the Arizona KidsCare program. Along the way, however, Hull and her colleagues made several important compromises.

First, the eligibility criterion for KidsCare is less generous than initially proposed. The goal now is to phase in coverage over a two-year period to about 72,000 children.³⁵ During phase one (November 1998 until October 1999), the program will cover children in families with income less than 150 percent of the FPL. The income criterion goes up to 200 percent in phase two (October 1999). Children in families with income above 150 percent will be required to pay premiums.

Second, the state decided to create a separate state CHIP rather than using CHIP dollars to finance a Medicaid expansion. This option provided the state with greater discretion to set program policy. At the same time, however, the state agreed that AHCCCS would administer nearly all of the CHIP initiative. There is a single CHIP/Medicaid application form. Only clients who are not Medicaid eligible are considered for CHIP.³⁶ Most CHIP beneficiaries enroll in the AHCCCS managed care network. Managed care plans provide both groups with the same benefit package and receive the same capitation rate.

Third, the legislature agreed that not all KidsCare beneficiaries had to participate in AHCCCS. Instead, parents could choose to enroll their children in the direct services program or they could sign them up with a health plan in the state employee health insurance plan. Under the direct services program, children can receive free care from any participating provider and the provider is paid on a fee-for-service basis by the state. The goal is to provide an option for families who are opposed to enrolling in managed care.³⁷

With these various compromises in place, the initiative received federal approval in September 1998 and the state began implementation in November 1998. By all accounts, however, enrollment is much lower than expected. The state expected that more than 20,000 youngsters would sign up during the first few months; as of February 1999, there were fewer than 10,000 CHIP enrollees.

State officials suggest three reasons for the lower-than-expected enrollment. First, nearly 36 percent of all CHIP applicants end up on Medicaid. Second, many potential applicants are unaware of the program. Third, some potential applicants are deterred by the administrative burden of enrollment. In an effort to overcome these obstacles, the state recently developed a simplified KidsCare application. State officials also have decided to accept applications by mail. Finally, the state has just hired a marketing agency to conduct a one-year, $700,000 advertising campaign.

State officials also note that nearly all of the CHIP enrollees have joined the AHCCCS managed care system. The main reason is that the benefit package in the direct services program is less generous than the managed care alternative. For example, neither behavioral health nor emergency room care is covered in the direct services program. At the same time, there is little indication that health care providers are encouraging clients to choose the direct care option, nor is the federal government likely to provide federal funding for the initiative. For all of these reasons, there seems to be little chance that the program will survive without significant alteration.

Arizona's Efforts to Increase the Availability and Affordability of Private Insurance

Over the past decade, nearly every state has enacted a series of efforts to make health insurance more available and more affordable. These initiatives focus on three problems in the health care system. First, employers in the small-business community often cannot afford to provide health insurance to their employees. Second, the employees in these companies generally earn too little to purchase their own health insurance policies. Third, persons with high-cost medical needs are often excluded from the individual insurance market even if they can afford a relatively high premium.

Arizona lags behind most states in its effort to reform the small-group insurance market. To be sure, the state does require insurers in the small-group (2 to 50 employees) market to issue policies to all small groups that apply. State law also permits insurers to offer "bare bones" (and thus less expensive) insurance policies to those in the small-group market. By most accounts, however, neither the guaranteed issue law nor the bare bones insurance program has had much of an impact on the small-group market.

There are, nonetheless, two state initiatives that deserve special attention. The first is a state-administered program that enables small businesses to buy health insurance from health plans with AHCCCS contracts. This program, administered by a state agency known as the Health Care Group of Arizona, is in the midst of a financial crisis and may not survive. The second initiative is called the Premium Sharing Program; it also is administered by the Health Care Group and subsidizes the cost of health insurance for persons with income below 200 percent of the FPL (400 percent for the chronically ill).

Health Care Group of Arizona

The original AHCCCS legislation in 1981 authorized small employers to buy health insurance from health plans with AHCCCS contracts. The goal was to provide a low-cost insurance alternative for the small-business community. Despite the legislative intent, however, state officials delayed for years the implementation of the small-business initiative. One problem was that AHCCCS officials had little interest or investment in the program. A second problem was that the state did not appropriate any funds for the initiative. The costs of the program (both start-up and ongoing) were to be borne entirely by subscriber premiums.

To overcome these obstacles, program supporters eventually persuaded the legislature to create a new agency, the Health Care Group, and to place the small business initiative under its jurisdiction. The Health Care Group obtained more than $700,000 in grant funds from The Robert Wood Johnson Foundation, which enabled the program to get started. At the same time, state officials decided to charge participating health plans a $4 per member per month fee to cover administrative costs. With these changes in place, program implementation began in early 1988.

The rules of the new program were rather straightforward. First, participating health plans were required to accept all applicants, regardless of health status, so long as the applicant worked full-time in a firm with 1 to 50 employees. Second, the plans were required to charge a modified community rate: they could vary rates on the basis of age and county of residence but not on the basis of health status. Third, the plans could not raise rates by more than the medical component of the metropolitan Phoenix consumer price index (CPI). Fourth, the state would not subsidize the cost of the program. State officials expected the health plans in the program to survive financially on the premiums paid by subscribers.

During the early 1990s, the program grew at an incremental pace. There was some consumer interest, though less than expected. State officials hoped to have 50,000 subscribers by early 1997, but fewer than 22,000 had signed up as of that date. The main problem was that the four participating health plans were all losing money on the initiative and therefore had little incentive to engage in a major marketing campaign. In 1997, for example, the four plans together lost $7.6 million on their Health Care Group product.³⁸

By mid-1997, the four health plans were threatening to exit the program. Their main complaint was that they were subsidizing what had evolved into a state-administered high-risk pool. State officials did not dispute the charge. The state itself issued a report that showed that the premiums paid by healthy subscribers were 30 percent higher than the premiums available in the regular commercial market, while the premiums paid by high-risk enrollees were 30 percent lower than what was available on the open market. Not surprisingly, employers were enrolling their healthy employees in commercial plans and their high-risk employees in the Health Care Group.

In the summer of 1998, in response to a growing crisis, the state legislature appropriated $8 million to the three health plans that agreed temporarily to stay in the market. At the same time, however, the legislature enacted four other provisions designed to save the program. First, the state lifted the requirement that the plans could not raise rates by more than the medical component of the metropolitan Phoenix CPI. Second, participating employers with five or fewer employees were required to enroll all members of the firm in the Health Care Group; firms with more than five employees had to include at least 80 percent of all employees. Third, state officials changed the definition of "full-time employee" from 20 hours per week to 32 hours per week. Fourth, the legislature created a task force to examine the long-term structure of the program. State policymakers recognize that the Health Care Group will not survive as a community-rated high-risk pool that operates without state funding.

In the aftermath of the legislative activity, the three health plans each increased premium levels.³⁹ The higher rates, new rules, and negative publicity caused program enrollment to decline to 14,656 as of October 1998.

Arizona Premium Sharing Program

The Premium Sharing Program is designed to subsidize health insurance for persons with income below 200 percent of the FPL (400 percent for the chronically ill) who are not eligible for AHCCCS or KidsCare. The target population is the parents of children who are enrolled in AHCCCS or KidsCare (though one-quarter of all enrollees are under the age of 18). The goal is to serve between 5,000 and 7,000 persons (though no more than 200 chronically ill persons with income above 400 percent of the FPL). The program is administered by the Health Care Group and is available to residents of four counties: Cochise, Maricopa, Pima, and Pinal. The cost of the program is covered primarily by a $20 million annual allocation from the state's tobacco tax, supplemented by beneficiary premiums and copayments. The average client pays a premium of $18 per month.⁴⁰

Until recently, program enrollment was well below expectations. Although state officials expected to reach the enrollment cap by the summer of 1998, actual enrollment as of November 1998 was only 2,604. In response to the disappointing results, state officials redesigned and simplified the mail-in application form, eased the administrative burden on clients (by extending from 6 to 12 months the time between recertifications), and implemented a new outreach and marketing initiative. The results are encouraging. As of July 1999, enrollment had increased to nearly 4,600.

The Medical Safety Net in Arizona: Coping with the Rising Numbers of Uninsured

Arizona's unusually large population of uninsured persons increases the importance of the state's medical safety net. The main sources of care for the uninsured (and for AHCCCS clients) are the public hospitals in Phoenix and Tucson, the 28 community health centers (with 80 sites) around the state,⁴¹ and the public health clinics in every county. In the early 1990s, however, the safety net seemed in danger of crumbling. In Phoenix, the county hospital was half empty and losing more than $30 million a year. Around the state, community health centers struggled with reduced Medicaid reimbursement and increased numbers of uninsured patients. Several clinics were forced to close. The growing crisis prompted calls for legislative and regulatory relief.

By the late 1990s, however, the crisis in the medical safety net seems to have eased. Several factors explain this surprising development. The electorate increased the state's tobacco tax, and much of the revenue generated has gone to safety net providers. AHCCCS has agreed (after resisting for years) to guarantee that community health centers receive cost-based reimbursement. AHCCCS also has agreed to ease the reserve requirements imposed on county-based managed care plans. A private management firm has brought fiscal stability to the public hospital in Phoenix. The net result of these changes is a medical safety net that seems stronger and more secure than it did just a few years ago.

The turnaround began in November 1994, when Arizona voters enacted Proposition 200, which increased the state's tobacco tax by 40 cents, thereby generating between $120 million and $125 million per year in additional revenue. The 1994 referendum also required that 70 percent of the tax revenue generated be spent on programs to aid the state's uninsured.⁴² Since that time, some of the tobacco tax revenue has funded insurance expansion initiatives, such as KidsCare and the Premium Sharing Program. At the same time, however, the state legislature has allocated much of the new revenue to efforts to support the medical safety net. Indeed, the tobacco tax now finances four initiatives that, although relatively small, provide new aid to the safety net (amounting to about $20 per uninsured person).

Three of the tobacco tax programs target primary health care clinics, while the fourth focuses on safety net hospitals. The first program, known as the Part A Initiative, finances clinic efforts to increase provider capacity. In 1997, this initiative provided $6 million to 26 health care clinics around the state. The second program, known as the Part B Initiative, provides funding to subsidize care provided to the uninsured. In 1997, this program allocated $5 million to nine primary care clinics. The third and most recent effort provides $2.5 million to build, expand, or renovate safety net clinics; six clinics (including five in rural communities) divided these dollars. Fourth, there is the Children's Hospital Program, known as the Part C Program, which provides $2.5 million annually to three safety net hospitals.

Each of these four programs requires the recipients of the funds to engage in communitywide efforts to encourage uninsured persons with income less than 200 percent of the FPL to use the health care services offered by the clinic or hospital. Each has played an important role in strengthening the state's medical safety net.

Another factor contributing to the strengthened safety net is the recent settlement between AHCCCS and the state's community health centers. This dispute dates back to legislation enacted by Congress in 1989 that required state Medicaid programs to pay cost-based reimbursement to federally qualified community health centers. Shortly thereafter, several community health centers submitted bills to AHCCCS to cover the difference between their costs and the reimbursement received from managed care plans. For years, the state resisted paying these so-called wraparound payments. In October 1997, however, AHCCCS officials agreed to pay health centers $1.75 per member per month to cover the cost of the wraparound requirement.⁴³ AHCCCS also agreed to compensate several clinics for services previously rendered (three of the state's largest health centers received more than $2 million in such payments).

The third component in the safety net turnaround is the improved performance by the public hospital in Maricopa County. In the mid-1990s, this facility was underutilized, inefficiently administered, and losing more than $30 million per year. In response to the crisis, county officials tried to sell the hospital (and several public health clinics) to a California-based company called Healthcare Providers Inc. (HPI). This privatization initiative did not succeed.⁴⁴ Shortly thereafter, however, the county hired Quorom Health Resources to manage the hospital, along with the rest of the county health care system. By all accounts, Quorom installed a skilled management team that has engineered a financial turnaround for the institution.

Finally, AHCCCS exempts county-owned managed care plans, such as the Maricopa Integrated Health System, from the capitalization and reserve requirements imposed on other plans, as long as the county itself guarantees that it will cover any outstanding debts. This provision makes it far easier for the publicly funded health plans to remain in the market. In Maricopa County, for example, health plans with AHCCCS contracts generally are required to have $2.3 million in reserves; the Maricopa Integrated Health System is exempt from this requirement.

Even with these various initiatives, however, the future of the state's medical safety net is uncertain. One problem is the tenuous nature of the recent financial and managerial improvements. The tobacco tax revenue could be targeted elsewhere. The requirement that health centers receive cost-based reimbursement is being phased out. Fewer dollars are available under a federal program that provides supplemental payments to hospitals that treat a disproportionate share of the state's indigent population (the state's share was reduced from $96.6 million annually to $81 million).⁴⁵

A second and more fundamental problem, however, is the state's large and rising number of uninsured. With roughly 28 percent of the population uninsured (the highest percentage in the nation), the pressure on safety net facilities is bound to increase. Managerial efficiencies in the Maricopa system and elsewhere surely help, but these alone are not likely to provide long-term solutions. Persons without health insurance eventually need health care, and, more often than not, such care is provided by the safety net. The high cost of uncompensated care could lead to a renewed crisis in the state's medical safety net.

The Arizona Long-Term Care System

When Arizona enacted AHCCCS in 1982, it excluded long-term care services from the covered benefit package. AHCCCS paid for the acute care services received by indigent nursing home residents, but the counties remained responsible for paying the nursing home bill. State officials were unwilling to share in the cost of this sector of the health care system.

The exclusion of long-term care generated significant opposition. Counties complained about the budget pressure imposed by a county-funded long-term care system. Nursing home owners complained about variation in county coverage and about inadequate rates of reimbursement. Consumer advocates complained that clients were going without needed care, both in institutions and in the community. Even state officials acknowledged that the infusion of federal dollars that AHCCCS coverage would generate would produce improved care for many.

In late 1988, in response to these concerns, the Arizona legislature decided to provide long-term care benefits to AHCCCS beneficiaries, and the federal government approved the change to the state's 1115 waiver. There were, however, three unusual features to the long-term care program. First, the state required that beneficiaries receive long-term care services through managed care companies. Arizona is the only state in the nation with such a requirement. Second, the legislature created a separate managed care system for persons in need of long-term care. The new system, called the Arizona Long-Term Care System (ALTCS), provides a full range of medical services, from acute to long-term, to persons in need of long-term care.46 Third, the state required county governments to pay ALTCS costs not paid by the federal government. State policymakers were willing to bring federal Medicaid dollars into the long-term care system but were unwilling to contribute additional state dollars.

Who Is Eligible to Receive ALTCS Benefits?

As of September 1998, 25,331 persons were enrolled in ALTCS: 15,970 elderly or physically disabled people and 9,361 developmentally disabled people. Each of these beneficiaries satisfied the two basic criteria for participation in ALTCS. First, beneficiaries must meet the Medicaid financial eligibility criteria. (Applicants in institutions may have income up to 300 percent of the federal Supplemental Security Income [SSI] cash assistance criteria.)⁴⁷ Second, the beneficiary must be sufficiently disabled to be at risk for more than three months of services at a nursing home or an intermediate care facility for the mentally retarded.⁴⁸

On its face, the ALTCS eligibility criteria are similar to criteria used in most state Medicaid home- and community-based services waivers. So is the actual application process: ALTCS eligibility workers review income and assets while state-employed nurses and social workers review the applicant's functional disability. By all accounts, ALTCS interprets and implements the functional disability requirement more restrictively than do systems elsewhere. In Oregon, for example, the functional disability requirement disqualifies fewer than 1 percent of all applicants; in Arizona, the requirement disqualifies between 15 percent and 20 percent of applicants.⁴⁹ Many people are deterred from even applying. Still others lose their eligibility during the annual recertification process; as their health status improves, their eligibility for ALTCS comes to an end.

One consequence of the restrictive criteria is that ALTCS beneficiaries are more disabled than their counterparts in other states. According to one study, 96.7 percent of ALTCS clients in nursing homes are dependent in toileting and eating, compared with 65.4 percent of nursing home clients around the country.⁵⁰ The restrictive eligibility criteria also lead to complaints that needy persons are wrongfully denied benefits. This problem is especially troublesome for ALTCS clients who lose eligibility during the recertification process. These clients often have become dependent on a certain amount of long-term care services that they can no longer receive.

State officials suggest that the strict eligibility criteria enable them to concentrate scarce resources on those most in need. These officials also argue that less-restrictive criteria would encourage the so-called "woodwork effect": Less-disabled persons who could manage without paid long-term care would obtain such services (especially home- and community-based services) simply because the services were available.

At the same time, however, state officials did develop a transitional eligibility program in 1995 for persons who lose benefits during the recertification process. Under the transitional eligibility program, clients can receive six months of home- and community-based services even after a determination of functional ineligibility. This program enables clients who are discharged from nursing facilities to have a smooth transition back into the community. The program also enables clients to ratchet back their use of home- and community-based services. As of August 1998, 2,421 clients were in the transitional eligibility program.

There is an ongoing debate over needy persons who slip through the ALTCS safety net and go without needed long-term care services. State officials insist that nearly all of the truly needy are covered and that the others can function without government aid. Consumer advocates challenge this assumption and suggest instead that thousands of the poor go without needed services because of the tight eligibility criteria. So far, neither side has produced persuasive evidence. This information gap makes it impossible to evaluate the actual size of the "needy but not covered" group.

Which Health Plans Participate in ALTCS?

Eight health plans participate in the ALTCS program. Seven serve the aged and physically disabled, while one serves the developmentally disabled. Each of the seven plans that serve the aged and physically disabled has an exclusive contract to serve beneficiaries in one or more of the counties around the state. The eighth plan, which is operated by the state's Department of Economic Security (DES), serves all of the developmentally disabled around the state. Table 5 lists the eight plans and their enrollment as of September 1998. As the table makes clear, the vast majority of aged and physically disabled enrollees are enrolled in plans administered by county governments—the only noncounty systems are Ventana and Arizona Physicians IPA.

Table 5  Health Plans with ALTCS Contracts Plan Number of Enrollees Department of Economic Security 9,416 Maricopa Managed Care Systems 8,903 Pima Health Systems 2,530 Ventana Health Plan 1,175 Yavapai County Long-Term Care 712 Arizona Physicians IPA 608 Pinal County Long-Term Care 535 Cochise County Department of Health Services 488 Source: Materials Received from AHCCCS, 1999.

Each of these health plans receives a per member per month capitation rate from the state that is set by a combination of competitive bidding and negotiation. The average rate paid in 1998 for aged and physically disabled beneficiaries was $2,192 per member per month. The average rate paid for the developmentally disabled was $2,082 per member per month. In exchange for this reimbursement, the health plan is responsible for managing the care of the enrollee and ensuring that he or she has adequate access to the full range of primary, acute, and long-term care services.

An important characteristic of the current system is that it is not a competitive environment. Beneficiaries do not select a health plan; the developmentally disabled are enrolled in the DES plan, and the aged and physically disabled are enrolled in the plan with the contract in their county. Similarly, health care providers do not negotiate with different health plans; providers that do not contract with the single ALTCS health plan in an area are not able to serve the ALTCS beneficiaries. Finally, in the state's five largest counties, private-sector managed care organizations cannot bid for the ALTCS contracts to serve the aged and physically disabled. In these areas, county governments are guaranteed the ALTCS contract (these cover nearly 90 percent of all aged and physically disabled enrollees).

State officials want to infuse competition into the ALTCS system. For this reason, in 2001, state officials will eliminate the requirement that the county governments in Maricopa (i.e., Phoenix) and Pima (i.e., Tucson) administer the ALTCS plans in their counties. State officials also will eliminate the rule that Cochise, Pinal, and Yavapai Counties have the right of first refusal for the administration of the plans in their regions. Under the new rule, there will be open competition for the ALTCS contracts in all of the state's 15 counties.

The effect of the new rule is hard to predict. Private firms will surely bid for contracts in counties such as Maricopa. Two likely candidates are Ventana Health Plan, which already has ALTCS contracts in seven rural counties, and Arizona Physicians IPA, which has contracts in three counties. Other private firms are likely to bid as well. At the same time, local officials will work hard to retain the contracts, because ALTCS is too lucrative to voluntarily abandon. In Maricopa, for example, the ALTCS plan reportedly subsidizes losses incurred by the county's AHCCCS health plan. In this environment, the competition for the ALTCS contracts could become quite fierce. The competition could also generate significant political controversy, especially if the state decides to end the contracts with Maricopa and Pima Counties.

One way to minimize the conflict would be to allow more than one contractor to operate in a single county. Such a decision would also provide consumers with more choice and providers with more negotiating leverage. As of early 1999, however, state officials had yet to adopt a multiplan option.

The ALTCS Delivery System

Once clients are accepted into the ALTCS system, their health plan assigns a case manager to develop a long-term care service plan. The case manager and the client (and often the extended family) seek to provide care in the least-restrictive environment. For the least disabled, this means various home care services, ranging from home health care to home-delivered meals to homemaker services. For those unable to stay at home, there is a relatively small network of community-based residential providers, such as assisted living facilities. Finally, for the most severely disabled, there are nursing homes.

The Nursing Home Industry in Arizona

Compared with other states, Arizona has a relatively small system of nursing facilities. The number of nursing home beds per 1,000 state residents over the age of 65 is 27.1; the national average is 49.1.⁵¹ The occupancy levels of those beds, now around 86 percent, ranks 34th in the nation.⁵² There are 23.1 nursing home residents per 1,000 Arizona residents over the age of 65; the national average is 43.7.⁵³

Interestingly, however, the state's nursing home infrastructure is growing more rapidly than that of any other state in the nation. In 1980, for example, the state had 76 nursing homes with 6,197 beds. By 1995, there were 158 facilities with 17,264 beds. By 1998, the numbers were 171 facilities with 19,020 beds. The growth is unmatched by any other state. As a result, the number of Arizona nursing home residents increased by 9.1 percent between 1995 and 1996. During that period, the number of nursing home residents nationally declined by 0.2 percent.⁵⁴ If these patterns continue, Arizona's ratio of nursing home beds will soon surpass the national average.

Why is Arizona increasing its nursing home bed supply at the same time the rest of the country is moving in the opposite direction? The best explanation is rooted in Medicaid. In most states, Medicaid coverage of long-term care services began in the late 1960s and early 1970s. This coverage prompted a dramatic increase in nursing home beds. For the first time, there was a third-party payer for most nursing home care. As a result, the nation's supply of nursing home beds increased from 300,000 in 1963 to more than 1.3 million in 1977. By the early 1990s, many states were imposing caps on the number of licensed nursing home beds in order to minimize new growth. Even states without such moratoriums often required nursing home operators to navigate certificate-of-need requirements before developing new nursing home beds.

The pattern in Arizona is quite different. Arizona did not enact its Medicaid program until 1982, and it did not cover long-term care services until 1989. To be sure, county governments paid for the nursing home care received by some welfare beneficiaries, and Medicare and private insurance paid for a few others. Nonetheless, without Medicaid, comparatively few persons could afford the high cost of nursing home coverage, and there was little demand for new nursing home beds. Since 1989, however, the availability of Medicaid coverage has led to the kind of growth that other states experienced 20 years earlier. Moreover, state officials have not enacted the barriers to growth often found elsewhere; nursing home owners do not have to comply with a certificate-of-need process, and there is no cap on total nursing home beds.

There is, to be sure, significant conflict between the nursing home industry and the managed care industry. The nursing home owners, who now receive 70 percent of their revenue from ALTCS, complain that the per diem rates paid by the health plans are too low. These facilities would prefer to see the state award more than one contract in a particular county, thereby providing the homes with some negotiating leverage. The owners also complain that the health plan focus on home- and community-based care skews the case management process and discourages some beneficiaries from receiving needed institutional care. Despite these complaints, however, the state's nursing home industry seems to be thriving, primarily as a result of the enactment of ALTCS.

Home- and Community-Based Services in Arizona

There is an ongoing debate over whether making home- and community-based services available to Medicaid beneficiaries raises or lowers overall costs. All agree that home- and community-based services are usually cheaper than institutional costs on a per person basis.⁵⁵ Nonetheless, the availability of home- and community-based services might attract new beneficiaries who would not otherwise be placed in nursing homes. If there are enough of these new beneficiaries, overall costs will rise even while per capita costs decline. This is the so-called "woodwork effect." Federal officials, worried about the woodwork effect, have imposed a cap on the percentage of aged and physically disabled beneficiaries that can receive home- and community-based services. The goal is to encourage the state to cover only the most severely disabled persons.⁵⁶

Arizona officials argue that the cap is unnecessary because the state's rigorous preadmission screening processes deny ALTCS eligibility to all but the most highly disabled. State officials insist that to treat this population in the community is cost-effective. Two studies by Weissert and colleagues support this claim.57 Federal officials, while not abandoning the cap, have increased it from 5 percent of total institutional beneficiaries in 1989 to 50 percent in 1999.

As the cap has increased, ALTCS officials have encouraged program contractors to increase their reliance on home- and community-based services. The chief incentive is financial. The capitation rate that is paid to the contractors is based on an assumption about the level of home- and community-based services that the contractor will provide. For example, if the state assumes that the contractor will serve 40 percent of its clients in home- and community-based settings, it will set a rate based on that assumption. If the contractor serves more than 40 percent in these settings, it is entitled to keep roughly three-quarters of the savings. If the plan serves fewer than 40 percent in such settings, the state will cover only one-quarter of the loss (the extra nursing home costs).

With this sort of financial incentive, the percentage of aged and physically disabled beneficiaries receiving home- and community-based services has increased from 7 percent in 1989 to 41 percent in 1998. Table 6 shows the steady pace of the expansion.

Table 6   Percentage of ALTCS Clients Receiving Home- and Community-Based Services Year Percentage 1989 7 1990 14 1991 17 1992 16 1993 26 1994 30 1995 33 1996 36 1997 49 1998 41 Source: Materials received from AHCCCS, 1999.

By mid-1998, approximately 5,800 ALTCS aged and physically disabled beneficiaries were receiving home- and community-based services. The vast majority of these clients (roughly 4,900) receive care in their own homes. The most common in-home services are home health care, home-delivered meals, and homemaker services. The other 900 beneficiaries live in community-based residences. The community-based residential options are assisted living facilities, adult foster care homes, behavioral health homes, and centers for the traumatically brain injured. Table 7 lists the various residential options and the number of persons enrolled in each.

The state hopes to increase the number of beneficiaries living in assisted living facilities. This is a new goal for state policymakers. Before 1998, state officials were unsure of the cost-effectiveness of the assisted living approach and capped the number of beneficiaries that could reside in such facilities.58 An evaluation of the ALTCS assisted living program demonstrated high client satisfaction combined with significant cost savings, prompting the legislature to eliminate the caps.

Ironically, however, the main obstacle to growth in the assisted living industry is the ALTCS reimbursement rates. Most assisted living facilities will not accept the rate paid by the ALTCS program contractors, and sufficient demand still exists in the higher-paying commercial market to eliminate the need to accept the lower-paying Medicaid rate. As a result, those few facilities that accept Medicaid clients have waiting lists. At the same time, however, the program contractors have little incentive to increase the rates. Instead, these contractors have an incentive to provide clients with in-home services, which are generally far less costly than assisted living facilities.

Table 7  Services Received by Aged and Physically Disabled ALTCS Beneficiaries Residential Option Persons Enrolled In-home services 4,900 Adult foster care 530 Assisted living homes (10 or fewer units) 155 Assisted living centers (11 or more units) 151 Behavioral health homes 60 Home for the traumatically brain injured 6 Source: Materials received from AHCCCS, 1999.

Paying the ALTCS Bill: Controversy over the State Share

The original ALTCS legislation required the county governments to pay the nonfederal share of the ALTCS bill. Perhaps surprisingly, however, the legislation did not simply require each county to pay for the cost of its own residents. Instead, the county share was based on its share of the statewide long-term care bill in 1988, the year before the implementation of ALTCS.

Over the next several years, county governments lobbied for legislative relief from the growing burden of the long-term care bill. The counties all argued that the state should pay at least some of the ALTCS bill. Moreover, the counties with relatively slow growth in their long-term care programs challenged the inequity of basing the county share on 1988 spending patterns. In effect, the slow-growth counties were subsidizing some of the costs incurred in higher-growth regions. Before 1997, however, the state legislature was unwilling to alter the ALTCS payment formula. The legislature would occasionally allocate special funds to counties in fiscal distress, but such relief was short term, inconsistent, and (according to the counties) inadequate.

The counties won more systematic relief in November 1997, when the legislature revised the long-term care payment formula. The new system requires the state to pay 50 percent of the nonfederal share of any amounts above what the counties actually paid in state fiscal year 1997-98. The legislation is hardly a sweeping county victory—the counties had lobbied for a 50/50 split of the ALTCS bill. Nonetheless, the legislature is unlikely to revisit the issue anytime soon.

Links between ALTCS and Medicare

Arizona policymakers believe that persons enrolled in both Medicaid and Medicare (the so-called "dual eligibles") too often receive unmanaged care. The lack of coordination is due to the different rules imposed by the two programs. AHCCCS requires beneficiaries to belong to a managed care company with an AHCCCS contract. Medicare allows beneficiaries to either enroll in a health plan with a Medicare contract or stay in the traditional fee-for-service system. Large numbers of the dual eligibles are therefore enrolled in two separate delivery systems: the AHCCCS managed care system and one of the two Medicare systems.

The problem of managed care and the dual eligibles is not limited to Arizona. The percentage of Medicare beneficiaries in Arizona that also are Medicaid eligible is 7.7 percent, far lower than the 13.1 percent national average.59 State officials, however, have developed an aggressive program to encourage dual eligibles to use their AHCCCS providers when accessing Medicare services. The key strategy is to pay the Medicare cost sharing (copayments and deductibles) when clients receive in-network care but to refuse to pay these bills for clients who obtain out-of-network care.

The current policy emerged from a more ambitious proposal, in the mid-1990s, to integrate the Medicaid and Medicare delivery systems. Under that proposal, ALTCS clients who did not explicitly choose to stay in the traditional Medicare system would receive all Medicare-covered services from their ALTCS contractor. The negotiations over the Arizona proposal lasted for more than two years. There were three key disagreements between state and federal officials. First, federal officials rejected the proposal that clients who did not choose would be automatically defaulted into the single integrated delivery system. Second, federal officials pressed the state to pay the Medicare copayments even when clients received out-of-network care. Third, state and federal officials could not agree on the rate AHCCCS would receive for providing the Medicare services. Unable to resolve these differences, the state eventually withdrew its proposal and enacted its current policy.

Evaluating the ALTCS System: Literature on Cost and Quality

McCall and colleagues have completed the most comprehensive study of the ALTCS system.⁶⁰ One goal of the study was to evaluate the quality of care received by ALTCS beneficiaries. The strategy was to review and compare the medical records of nursing home patients in Arizona and New Mexico during 1991 and 1992. The review focused on particular medical indicators, such as the incidence of pressure sores and the number of falls and fractures. The findings suggested that Medicaid beneficiaries in New Mexico generally received higher-quality nursing home care than did their Arizona counterparts. Specifically,

. . . nursing home residents in the ALTCS program were more likely to experience a decubitus ulcer, a fever, and a catheter insertion than nursing home residents served by the New Mexico Medicaid program.⁶¹

As McCall noted, however, the findings must be viewed with caution. First, the researchers relied on data generated in 1991 and 1992, just a year or two after ALTCS was first implemented. State officials suggest that it took longer for managed care to significantly affect the quality of care. Second, nursing homes in both Arizona and New Mexico are reimbursed on a per diem basis. ALTCS contractors have not sought to capitate either the nursing home industry or the long-term care industry in general. From the provider's perspective, therefore, the Medicaid programs in Arizona and New Mexico are not particularly different.

The McCall study also examined whether Arizona had saved money by adopting ALTCS instead of a traditional Medicaid program. On this issue, the researchers found a major impact. According to their analysis, between 1989 and 1993, Arizona saved 18 percent on medical services (on behalf of persons in the long-term care system) and 16 percent overall by using the ALTCS model instead of a traditional Medicaid program. Moreover, the savings had increased dramatically over time, moving from 0.2 percent in 1989 to 8 percent in 1990 to 21 percent in 1992 and 1993. The three key reasons for the cost savings were (1) the state's impressive (and not inexpensive) administrative infrastructure; (2) the preadmission screening process, which limits access to the program; and (3) the incentives in the capitation rate to encourage home- and community-based services.

Weissert and colleagues also presented evidence to suggest that Arizona had saved money by using the ALTCS model instead of a traditional Medicaid program.62 Again, the key was the state's ability to encourage beneficiaries to use home- and community-based services rather than nursing home placement. Weissert asserts that the amount saved was at least $3 million during the two-year period from 1996 to 1997. His argument is strengthened by data that show that Arizona spends less (22.3 percent) of its Medicaid bill on long-term care than any other state (the national average is 35 percent).⁶³

However, two important caveats bear mentioning. First, there is an ongoing debate over whether the preadmission screening process is too restrictive and denies eligibility to needy persons. Persons denied needed long-term care services may end up needing higher levels of acute and other health care services or may experience a lower quality of life. Second, there needs to be further study of the quality of care provided in the AHCCCS system. The McCall study suggested some areas of concern: Medicaid beneficiaries in New Mexico seemed to get better nursing home care than did their Arizona counterparts. Nonetheless, that study is relatively dated, looked at only one other state, and did not take into account the recent emphasis on home- and community-based services.

The Arizona Behavioral Health System

The behavioral health system available to low-income Arizona residents is in the midst of a crisis. There are two main problems. First, the federal government recently terminated its accreditation of the state's only mental hospital, making it ineligible to receive Medicare and Medicaid funding. By all accounts, conditions in the facility have only worsened since the loss of accreditation and federal funding. Second, the state recently terminated the contract of the largest behavioral health plan in its managed care network. The termination caused significant service disruption for many enrollees. It also resulted in protracted litigation between the state and the health plan.

The story begins with litigation in the late 1980s challenging the conditions in the state's large mental facility. The consumer advocates who initiated the litigation hoped to force the state to discharge most of the institutionalized population to community-based residential settings. The court agreed that the state needed to begin the deinstitutionalization process. By the mid-1990s, hundreds of patients were released to community-based settings, and persons who previously would have entered the state hospital were referred elsewhere. In turn, the state reduced the funding allocated to the hospital and the hospital dramatically reduced the number of staff.

Unexpectedly, however, there has been a dramatic upswing in the number of persons admitted to the state mental hospital. The majority of the new patients are referred by the criminal justice system. Every month, for example, there are five or six referrals under the state's recent Sexually Violent Predator law.64 The patient population today is far more problematic and hard to treat than the population of a decade ago. At the same time, the staff-patient ratio is well below the norm, and many of the staff that remain are not trained in treating a forensic clientele. Given the deteriorating conditions, it is hardly surprising that a 1998 audit by the Joint Commission on the Accreditation of Health Care Organizations resulted in the loss of federal accreditation.

In response to the crisis, Governor Hull has announced that her number one health care priority for 1999 is to address the problems in the state mental hospital. The short-term plan is contained in legislation enacted in February 1999 that allocates $4.4 million to address staff shortages and overcrowding. More important, however, the governor hopes to persuade the state legislature to allocate funds from the tobacco settlement to build a new facility.

The deinstitutionalization movement also prompted a crisis in the community-based mental health system. This system was itself in the midst of a fundamental transition during the early 1990s. It was not until 1989 that the legislature authorized Medicaid coverage for behavioral health services, and the new behavioral health program had three unusual features. First, the state required that beneficiaries receive behavioral health services from managed care companies. Second, the legislature created a separate managed care system for persons in need of behavioral health care. Under the new system, there are five behavioral health regions, and in each region one health plan has a contract to provide behavioral health services to AHCCCS beneficiaries.⁶⁵ Third, the legislature required AHCCCS to delegate responsibility for the behavioral health managed care system to the state's Department of Health Services (DHS).

The largest of the five behavioral health regions is Maricopa County. In 1992, DHS awarded the behavioral health contract for the Maricopa region to ComCare. Over the next several years, however, state officials became increasingly dissatisfied with the contractor's performance. One area of concern was the contractor's oversight of patients discharged from the state mental hospital. Consumer advocates and others charged that ComCare had far too few case managers for its patient population and that it failed to pay for various needed services. As a result, in late 1997, DHS terminated the ComCare contract and hired a new contractor, ValueOptions. ComCare sued, challenging the termination of its contract. The litigation was eventually settled when the state agreed to pay $9.5 million to the plan; however, the controversy over the adequacy of the community-based mental health residences continues.

Conclusion

The health insurance system in Arizona is in the midst of a crisis. At the beginning of the decade, roughly 21 percent of the state's population was without health insurance. By 1997, that number was up to 28 percent, the highest in the nation. Ironically, the insurance crisis is occurring despite the state's strong economy and low rate of unemployment. The best explanation is that the state's economy is dominated by service-sector jobs that are relatively low paying and often do not provide insurance or other benefits. Only two states, Arkansas and New Mexico, have a lower percentage of persons with employer-based health insurance coverage. Welfare reform and the concomitant decline in the number of Medicaid beneficiaries exacerbate the insurance crisis.

State officials have sought without much success to reverse the trend toward more uninsured persons. The most dramatic proposal is to liberalize Medicaid eligibility criteria for adults from 32 percent of the FPL up to 100 percent (although the expansion would be limited to 150,000 persons). Federal officials have so far rejected the proposal, both because it is not budget neutral and because of the proposed enrollment cap.

State officials have implemented three other programs designed to reduce the number of uninsured. The state's CHIP initiative, KidsCare, covers children at or below 150 percent of the FPL. The Health Care Group provides low-cost insurance to the small-business community. The Premium Sharing Program subsidizes insurance for low-income persons in four counties around the state. By all accounts, however, enrollment is below expectations in all three initiatives. The disappointing results are explained by several factors, including the administrative burden of the application process, the lack of outreach and marketing, and the premium-based nature of the programs. State officials are working hard to minimize these obstacles and to encourage increased participation.

State officials are also struggling to reform the state's behavioral health system. The main culprit is an admittedly inadequate state mental hospital. Governor Hull hopes to use tobacco settlement funds to build a new facility. At the same time, state officials are working to improve the performance of the behavioral health organizations that manage the community-based mental health services received by Medicaid beneficiaries. The state has terminated the contract of a large behavioral health organization and is monitoring more closely the performance of the remaining plans.

As they cope with these crises, state officials also face three other important challenges. First, fewer than 10 percent of all AHCCCS beneficiaries choose their own managed care plans; the rest are autoassigned. The unusually high autoassignment rate suggests major problems with the state's system of marketing and enrollment.

Second, commercial health plan participation in AHCCCS is on the decline. The health plans argue that rates are inadequate, especially in rural regions. The trend toward fewer health plans in the Medicaid market means that the state's renowned system of competitive bidding is at risk.

Third, the state's long-term care managed care system is about to undergo fundamental change. It is hard to overstate the importance of the state's decision to permit private contractors to bid for the ALTCS contracts in Maricopa, Pima, Cochise, Pinal, and Yavapai Counties. State officials still need to determine the parameters of the competition. For example, it is unclear whether more than one contractor will be permitted in a county. Under any scenario, however, the competition is likely to generate significant political controversy as well as uncertainty and flux in the state's long-term care delivery system. Managing this transition is sure to be an important priority for state health care officials.

---

Notes

1. Arizona Department of Commerce, "The Gold Sheet: Arizona Economic Development Update," November 1998.

2. JoAnn Lamphere, Normandy Brangan, Sharon Bee, Rafael Semansky, Reforming the Health Care System: State Profiles 1998 (American Association of Retired Persons 1998).

3. The state's unemployment rate of 3.7 percent is one of the lowest in the nation.

4. The Arizona Town Hall, "Meeting the Challenges and Opportunities of Arizona's Growing Senior Population," 1998.

5. Urban Institute analysis of Current Population Survey data.

6. Lamphere et al. 1998.

7. Ibid.

8. Urban Institute data; see table 1.

9. Arizona Health Care Cost Containment System, "Managed Care Demonstration for American Indians Residing on-Reservation: 1994-95," prepared for the Health Care Financing Administration (HCFA) (1994).

10. Urban Institute analysis of Current Population Survey data.

11. The deserving poor were those (such as aged, blind, and disabled persons) who were outside the workforce through no fault of their own.

12. James Fossett and Associates, "Managing Accountability in Medicaid Managed Care: The Politics of Public Management" (Nelson A. Rockefeller Institute of Government 1999).

13. As of 1996, 37.3 percent of nonelderly persons below the FPL in Arizona were covered by AHCCCS; the national figure is 46.8 percent (Lamphere et al. 1998).

14. Beginning in the late 1980s, Congress required the states to expand the eligibility of pregnant women and children. Since AHCCCS operates under an 1115 waiver, state officials could have claimed an exemption from these new federal rules. Instead, AHCCCS has complied with the eligibility expansion mandates. AHCCCS covers pregnant women and infants in families with income below 140 percent of the FPL, children age six and younger in families with income below 133 percent of the FPL, and children between ages 6 and 15 in families with income below 100 percent of the FPL.

15. Robert Pear, "Poor Women Lose Medicaid Coverage Despite Eligibility," New York Times, April 12, 1999, page 1.

16. Some policymakers speculate that the disparity is largely due to the long history of low rates of institutionalization for the mentally retarded and mentally ill. State officials argue that the savings are generated by the state's unique program of managed care.

17. Lamphere et al. 1998.

18. Nelda McCall, "The Arizona Health Care Cost Containment System: Thirteen Years of Managed Care in Medicaid," Kaiser Family Foundation, July 1996.

19. As discussed later in this report, the acute care managed care companies provide mental health services only to persons between ages 18 and 20 who are not considered to be seriously ill.

20. The 1997 Balanced Budget Act requires that beneficiaries have a 90-day opt-out period. Arizona officials argue that the state's 1115 waiver exempts it from this requirement.

21. State officials hire actuaries to develop rate bands for the different rate cells. Health plans that bid below the band minimum are raised to that amount. Plans that bid too high are given a final counteroffer, usually in the bottom quarter of the rate band.

22. John Holahan, Suresh Rangarajan, and Matthew Schirmer, "Medicaid Managed Care Payment Rates in 1998," Health Affairs, May/June 1999, pp. 217-227.

23. Ibid.

24. The Phoenix Health Plan acquired between 80 percent and 90 percent of the former Blue Cross enrollees.

25. Nelda McCall, C. William Wrightson, Jodi Korb, Michael Crane, William Weissert, John Wilkin, and Susan Haber, "Evaluation of Arizona's Health Care Cost Containment System Demonstration—Final Report," February 1996.

26. U.S. General Accounting Office, "Arizona Medicaid: Competition among Managed Care Plans Lowers Program Costs" (GAOI/HEHS-96-2, October 1995).

27. AHCCCS, "Survey Shows AHCCCS Members Are Happy with Their Care," Flinn Foundation, 1995.

28. Lamphere et al. 1998.

29. McCall et al. 1996.

30. GAO 1995.

31. In response to task force recommendations, state officials raised dental reimbursement by 26 percent, which has encouraged additional dentists to participate and has increased dental utilization among children by 10 percent.

32. James Fossett and Associates 1999.

33. The program appropriates an additional $19.3 billion for the following five years.

34. Howard Fischer, "GOP Effort Seeks to Oust Hull in Primary: Kids' Health Proposal Upsets Conservatives," The Arizona Daily Star, April 21, 1998; Chris Moeser, "Capitol Games Just Get Weirder: Dems, Not GOP, Support Hull on 'KidsCare' Plan," The Arizona Republic, April 21, 1998.

35. The original proposal would have had a much shorter phase-in period.

36. CHIP enrollees also must not have had health insurance (other than Medicaid) for the six months prior to their application.

37. Arizona is funding the direct services program without any federal aid. The explanation is that federal law prohibits states from using more than 10 percent of CHIP dollars for outreach and education or to directly subsidize the cost of care provided to children. As a result, the $8 million cost for the first year of the program is covered by state tobacco tax dollars.

38. Health Care Group of Arizona, "Summary Report of Health Care Group Technical Assistance Meeting," February 2, 1988.

39. The Arizona Physicians IPA raised premiums by 23 percent after the premium cap was lifted.

40. The size of the premium depends on the client's income: Persons with income below 200 percent of the FPL pay up to 4 percent of their income, while the chronically ill with income above 200 percent of the FPL pay $410 per month.

41. There are 11 federally qualified community health centers, along with 3 Indian Health Service clinics and 14 "look-alike" clinics.

42. Proposition 200 also required that 23 percent of the revenue be spent on health education and antitobacco marketing campaigns, 5 percent be spent on research aimed at preventing and treating tobacco-related disease, and 2 percent be spent on corrections.

43. In 1997, in the Balanced Budget Act, Congress declared that the cost-based reimbursement requirement would slowly be phased out. Arizona officials plan to phase out the wraparound payments at the same time that Congress phases out the cost-based reimbursement requirement.

44. There were both legal and financial obstacles to the sale. First, HPI refused to produce a $20 million bond that the state required as collateral. Second, legal services groups challenged the proposed sale, noting that the deed for the property contains a restriction requiring the hospital to remain publicly owned.

45. Governor Jane Hull, "Executive Budget Message—FY 1999 Budget."

46. ALTCS covers only services not paid for by some other insurer. For example, Medicare beneficiaries receive ALTCS coverage for only those services not covered by Medicare.

47. In 1998, the income cap was $1,482 per month for an individual.

48. The acute care plans are responsible for the first 90 days of long-term care services received by the beneficiary.

49. William G. Weissert, Timothy Lesnick, Melissa Musliner, and Kathleen A. Foley, "Cost Savings from Home and Community-Based Services: Arizona's Capitated Medicaid Long-Term Care Program," Journal of Health Politics, Policy and Law, December 1997.

50. Ibid.

51. Robert W. Bectel and Natalie Graves Tucker, "Across the States 1998: Profiles of Long-Term Care Systems" (American Association of Retired Persons 1998).

52. Ibid.

53. Ibid.

54. Ibid.

55. The exception to the rule is when an intensive package of home health services is provided to individuals with extensive care needs.

56. There is no cap on the number of developmentally disabled persons who can be served in home- and community-based settings. As of mid-1998, nearly 98 percent of these beneficiaries were served in noninstitutional settings.

57. Weissert et al. 1997; William Weissert, Michael Chernew, and Achamyeleh Gebremariam, "Cost-Effectiveness of Home and Community-Based Services in the Arizona Long-Term Care System—An Update," report to the U.S. Health Care Financing Administration, October 1998.

58. Under previous state rules, no more than 700 beneficiaries could live in what were known as supportive residential facilities (now referred to as assisted living centers), while up to 200 clients could live in adult care homes (now referred to as assisted living homes).

59. Kaiser Commission on Medicaid and the Uninsured, "State Facts: Health Needs and Medicaid Financing," February 1998.

60. The findings are summarized in McCall 1996.

61. McCall 1996, p. 16.

62. Weissert et al. 1997; 1998.

63. Kaiser Commission on Medicaid and the Uninsured 1998.

64. Under this law, persons convicted of sexually violent crimes are often referred to the state mental facility after they have served their prison sentences.

65. There are two exceptions to this approach. First, the acute care managed care companies provide mental health services to persons between the ages of 18 and 20 who are not considered to be seriously ill. This exception is intended to enable AHCCCS officials to evaluate whether traditional managed care plans should be allowed to deliver behavioral health services. Second, ALTCS beneficiaries receive all of their care, including their behavioral health care, from their ALTCS contractor.

---

About the Author

Michael Sparer is an associate professor at the Joseph L. Mailman School of Public Health at Columbia University. He studies and writes about the politics of health care, with an emphasis on the state and local role in the American health care system. Mr. Sparer is the author of Medicaid and the Limits of State Health Reform (Temple University Press, 1996) as well as numerous articles and book chapters.

---

Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact [email protected].

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.