• Contents
• Introduction
• Study 1: Factors Contributing to Variations in Per Capita State AIDS Drug Assistance Program Expenditures
• Study 2: ADAP Supplemental Funds
• Study 3: The Role of Title I-Funded AIDS Pharmacy Assistance Programs in Ensuring Access to HIV Therapeutics
• Study 4: Issues Associated With Capped Enrollment and Waiting Lists

Study 1
FACTORS CONTRIBUTING TO VARIATIONS IN PER CAPITA STATE AIDS DRUG ASSISTANCE PROGRAM EXPENDITURES

Jeffrey Levi, Julia Hidalgo, John Palen, E. Blaine Parrish, Kendra Williams, and Anthony S. Lara

Background

State and territorial AIDS Drug Assistance Programs (ADAPs) (referred to in this report as State ADAPs) report substantial variability in resources for the purchase of HIV medications. Some State ADAPs have access not only to Title II ADAP earmark and base funds but also Title I and State and local government funds. Other State ADAPs are entirely funded by ADAP earmark and Title II funds. Some State ADAPs reportedly have been successful at maximizing their funds through effective cost-containment measures such as rebate and pricing agreements, retrospective billing to Medicaid, and health insurance purchasing. State ADAPs also vary in their eligibility criteria and benefits, as well as the extent to which they coordinate with Medicaid or other payers to ensure that the Ryan White Comprehensive AIDS Resources Emergency (CARE) Act is the payer of last resort (PLR).

As a result of the significant variability in State ADAP programmatic design, the per capita expenditures made by State ADAPs are likely to range widely. Issues of equity have been raised among policy makers, consumers, and advocates. The extent to which per capita expenditures actually vary is unclear, however, as is the influence of factors associated with that variability. Further analysis is required to obtain this information and to identify policy measures that might be adopted to ensure parity in per capita expenditures while ensuring continued contributions to ADAPs by State and local governments and other funders.

The goal of this project was to inform the HRSA HIV/AIDS Bureau (HAB) regarding the extent to which per capita State ADAP expenditures vary and the role of various factors in that variation (See Table 1). This study considered the following factors:

• Source and amount of non-Federal ADAP funds
• State ADAP eligibility requirements
• State ADAP programmatic design, such as mechanisms used to purchase and distribute drugs
• Covered ADAP benefits, such as the drug formulary
• Coordination with other funders to ensure that CARE Act funds are used as the PLR
• Drug payment levels and cost-containment strategies such as negotiated price reductions and rebate agreements
• The external health insurance environment that may result in increased or decreased demand for ADAP enrollment.

Methodology

Data Sources

All data routinely gathered by the HAB ADAP staff for the period from Federal fiscal year (FY) 1999 through FY 2003 were reviewed, including the following data sources:

• State Title II ADAP grant applications for FY 2001 through FY 2003 and ADAP-At-A-Glance tables that summarize the ADAP State Profiles submitted in the grant applications
• ADAP Monthly Reports submitted by State ADAPs for January 2001 through June 2003
• ADAP Quarterly Pricing Reports submitted by State ADAPs for March 2001 through June 2003
• Calendar year (CY) 2002 ADAP reports submitted by State ADAPs in the CARE Act Data Reports (CADR)
• State population-based health insurance enrollment data from the U.S. Bureau of the Census Current Population Survey (CPS): Annual Social and Economic Supplement obtained from: http://www.kff.org/uninsured/ upload/29340_1.pdf and published by the Kaiser Commission on Medicaid and the Uninsured (Health Insurance Coverage in America 2002. Washington, DC: Kaiser Family Foundation, December 2003)
• Follow-up telephone discussions with State ADAP staff to address questions that arose during the course of the analysis.

George Washington University (GWU) staff analyzed ADAP data to assess timeliness, completeness, and accuracy. GWU staff also consulted with HAB and State ADAP staff to determine which ADAP databases should be used in the analyses shown in this report and how to interpret the results.

Computing Per Capita Expenditures

Annual ADAP State Profile data submitted by State ADAPs were used to calculate the per capita expenditures made by each State ADAP for FY 1999 through FY 2003. Expenditure data submitted by State ADAPs were actual expenditures for FY 1999 through FY 2002 and estimated expenditures for FY 2003. Total clients served by the State ADAP were used as the numerator and total expenditures from all sources were used as the denominator:

Total ADAP expenditures, regardless of the funding source, were summed for each year to calculate the denominator. Sources of expenditures included Title II ADAP earmark funds, Title II base funds, Title I funds allocated to the State ADAP, State ADAP matching funds, other State revenue, other local funds, and other sources of funds. State ADAP per capita annual expenditures were used as the unit of analysis. State ADAP expenditure data included purchases made for the following five types of services:

1. Medications (including the price of individual medications and the dispensing fee)
2. Health insurance premiums
3. Health insurance deductibles
4. Health insurance co-payments
5. Flexibility policy expenditures for adherence and monitoring related to ADAP clients.

Per capita expenditures were calculated for each State ADAP for a 3-year period covering FY 2001, FY 2002, and FY 2003. State ADAP eligibility, coverage, cost-containment, Medicaid program coordination, and other relevant programmatic characteristics were appended to the per capita expenditure data for each State ADAP and fiscal year studied. Additional health insurance coverage data from the CPS were obtained for CY 2002, the most recent calendar year available. In the absence of trend data, individual State CY 2002 CPS data were used as a constant variable for CY 2001 and CY 2003.

Limitations

Computations of per capita expenditures did not include adjustments for inflation for the period in which this analysis was conducted. Inflation is likely to have some impact on the “real dollars” available to State ADAPs to purchase medications and health insurance premiums, deductibles, and co-payments. Because of the absence of State ADAP data regarding medication-specific and health insurance–related expenditures, it was not possible to identify medical market inflation data that closely mirrors the unique and varied benefit packages supported by the State ADAPs.

Pricing data submitted to HAB could not be used for this analysis. The pricing data apparently include medication purchases as well as co-payments and deductibles in the pricing data for pharmaceuticals purchased by individual ADAPs.

GWU staff encountered several challenges in assessing the factors associated with differences in per capita expenditures. Specific data issues that were identified and explored with HAB staff included the following:

• Interviews with State ADAP staff indicated that they may interpret HAB data collection forms differently.
• Per capita ADAP expenditure data available from HAB most directly measure medication purchasing and not the other services funded with ADAP dollars.
• Some State ADAPs have not submitted all required forms, have not completed sections of some forms, or have submitted their “best guesses.” For example, 28 percent of State ADAPs did not submit total expenditure data in their CY 2002 ADAP CADRs. As a result, GWU staff could not use CADR data as part of this analysis.
• Person-based expenditure data are not available to HAB for accurately computing the actual expenditures for the various types of services purchased. Despite substantial variability in the length of enrollment by clients in State ADAPs, the ADAPs do not report the data to HAB. Person-based data also are not available to differentiate among per capita expenditures made for deductibles, co-payments, health insurance premiums, or medication purchases.
• Because of the lack of available data, GWU cannot assess the contribution of carryover ADAP earmark and Title II base funds to variable per capita expenditures. Several State ADAPs are reported by HAB staff to have carried over large amounts of ADAP earmark funds but it is unclear what impact those funds have had on per capita ADAP expenditures.
• Incomplete pricing data hinders our ability to assess the direct relationship between pricing and per capita expenditures. In reviewing pricing data with HAB staff, it was concluded that pricing data were included in the price of co-payments and deductibles as well as in the actual average price of the medications purchased by State ADAPs. Variability in pricing by State ADAPs is not adequately captured by current data collection mechanisms.
• Given these various challenges, only estimated per capita annual ADAP expenditures are feasible at this time.

Major Findings

Trends in ADAP Enrollment and Expenditures That Affect Per Capita Expenditures

Demand for ADAP services increased sharply during the late 1990s with the adoption of highly active antiretroviral therapy (HAART) by clinicians treating HIV-infected patients and with the growth in the number of HIV-infected patients without other pharmaceutical coverage. By FY 2003, a total of 151,304 clients were receiving one or more ADAP-funded services annually. Annual total enrollment varied greatly among State ADAPs. For example, an average of 2,855 clients enrolled in State ADAPs in FY 2003, but enrollment ranged from 8 clients in Guam to 24,534 clients in California. State ADAPs reported that variability in enrollment significantly influences their ability to negotiate rebates and effectively use other cost-containment mechanisms.

The average number of clients enrolled in State ADAPs rose from 2,049 in FY 1999 to 2,922 in FY 2002. Average enrollment then dropped to 2,855 clients in FY 2003. The rate of change in average annual State ADAP enrollment between fiscal years was 18.3 percent between FY 1999 and FY 2000; 8.7 percent between FY 2000 and FY 2001; 10.8 percent between FY 2001 and FY 2002; and -2.3 percent between FY 2002 and FY 2003. The plateau of State ADAP enrollment may be attributable to several factors, including a flattening in the rate of newly identified HIV-infected individuals seeking enrollment, capping of State ADAP enrollment, and implementation of waiting lists because of inadequate funding.

Several State ADAPs account for the most enrollees. Ten State ADAPs accounted for 73 percent of total enrollment in FY 2003: California accounted for 17 percent of enrollment; New York had 15 percent of enrollment; Florida, 11 percent; Texas, 8 percent; Puerto Rico, 7 percent; New Jersey, 4 percent; Pennsylvania, 3 percent; Illinois, 3 percent, Georgia, 3 percent; and Washington, 2 percent. All other State ADAPs accounted for 27 percent of national ADAP enrollment.

State ADAP expenditures are based on ADAP earmark and Title II funds as well as other sources of CARE Act and non-Federal funding. National ADAP expenditures rose 10 percent between FY 1999 and FY 2000, 10 percent between FY 2000 and FY 2001, 11 percent between FY 2001 and FY 2002, and 9 percent between FY 2002 and FY 2003. By FY 2003, ADAP expenditures totaled more than $1 billion.

Title II earmark funds represented 66.7 percent of total ADAP funds expended. Title II supplemental funds accounted for 2.1 percent; Title II base funds, 2.3 percent; Title I funds, 2.2 percent; ADAP State matching funds, 0.5 percent; other State funds, 15.5 percent; and other funds, 10.8 percent. City and county revenue made up most of these other funds. State ADAPs vary substantially in the array of funding streams used to support their programs. Of the 57 State ADAPs funded in FY 2003, 42 State ADAPs (73.7 percent) did not receive Title I funds; 18 State ADAPs (31.6 percent) received no State non-matching funds; and 26 State ADAPs (45.6 percent) received no other funds, such as city or county revenue. All State ADAPs received ADAP earmarked funds.

Other State funds, excluding revenue used for the ADAP match requirement of the CARE Act, were an important source of State ADAP support between FY 1999 and FY 2003. Nonmatching State funds fluctuated slightly between FY 2001 and FY 2003: 47.6 percent, 45.6 percent, and 46.4 percent (respectively) of nonearmark State ADAP funds came from State revenue. Other funds, including city and county revenue, were also an important and growing source of State ADAP funds during this period. The percentage of State ADAP nonearmark funds derived from other sources rose from 27.8 percent in FY 2001 to 28.0 percent in FY 2002 and to 32.4 percent in FY 2003.

Following the trends in enrollment, 10 State ADAP expenditures accounted for 72 percent of total expenditures. New York, California, and Florida alone accounted for 45 percent of State ADAP expenditures. On average, ADAP expenditures rose per State from $13.4 million in FY 1999 to $19.2 million in FY 2003. The average expenditures across State ADAPs, however, obscure the substantial range in actual expenditures among State ADAPs. In FY 2003, for example, total expenditures per State ADAP varied from $91,319 to $200.8 million, reflecting differences in the amounts and sources of funds that support those programs.

Analysis of the trends in average State ADAP expenditures indicated relatively steady growth in average expenditures by State ADAPs between FY 1999 and FY 2002, with the rate of increase in expenditures averaging 9.9 percent between FY 1999 and FY 2000, 10.2 percent between FY 2000 and FY 2001, and 14.8 percent between FY 2001 and FY 2002. The trend in increased growth in expenditures changed between FY 2002 and FY 2003, with the increase in expenditures dropping to 3.3 percent. In actual dollars, the average increase in State ADAP expenditures was $1.3 million between FY 1999 and FY 2000, $1.5 million between FY 2000 and FY 2001, $2.4 million between FY 2001 and FY 2002, and $616,792 in FY 2003. A slight but steady increase was observed in the average expenditures of State ADAPs between FY 1999 and 2003. These underlying numeric distributional characteristics of expenditures, along with expenditure patterns among some State ADAPs, resulted in the difference between average expenditures among State ADAPs over time being not statistically significant.

Per Capita Expenditures

Trends were identified in State ADAP per capita expenditures among State ADAPs. Annual total per capita expenditures varied greatly among State ADAPs during the period studied. In FY 2003, for example, total per capita expenditures per State ADAP varied from $1,583 to $23,304. This variability reflects differences in the size of the enrollments in State ADAPs as well as in the amount and sources of funds supporting those programs. The median per capita expenditures, less likely to be influenced by outlier data, rose from $5,853 in FY 1999 to $6,446 in FY 2000 and $6,752 in FY 2001. Median per capita expenditures then decreased to $6,604 in FY 2002 and rose again to $7,099 in FY 2003.

Analysis of the trends in average State ADAP per capita expenditures indicated that average per capita expenditures by State ADAPs dropped 0.6 percent between FY 1999 and FY 2000 and then rose 1.5 percent between FY 2000 and FY 2001, 1.9 percent between FY 2001 and FY 2002, and 12.0 percent between FY 2001 and FY 2002. Reportedly inaccurate data submitted by one State ADAP skewed FY 2003 per capita expenditure data. When those data were removed from the analysis, the average per capita expenditure between FY 2002 and FY 2003 rose 7.5 percent.

In actual dollars, average State ADAP per capita expenditures dropped $40 between FY 1999 and FY 2000 and then rose $97 between FY 2000 and FY 2001, $126 between FY 2001 and FY 2002, and $493 in FY 2003 (deleting the inaccurate outlier for FY 2003). Using both parametric and nonparametric measures to assess differences between the mean and median per capita expenditures, no statistically significant differences were identified. Note that although no statistically significant differences in per capita expenditures were found, any incremental increase in expenditures has an impact on ADAP clients’ buying power.

Trends in per capita State ADAP expenditures were examined for the top 10 and bottom 10 States and territories for FYs 2001, 2002, and 2003. In FY 2003, for example, per capita expenditures were reported to be highest in Tennessee ($23,304), followed by Indiana ($13,120), Maryland ($12,020), Massachusetts ($11,695), Guam ($11,415), Michigan ($11,357), Ohio ($10,988), Delaware ($10,869), Maine ($10,765), and Oregon ($10,418). These States and territories vary substantially in enrollment, total funds expended, sources of ADAP funding (including total Title I funds allocated), eligibility requirements, formulary, and payment for insurance-related expenses. Note that the per capita expenditures reported in FY 2003 for Tennessee are likely to be in error as a result of data quality issues.

The States and territories with the lowest per capita ADAP expenditures were also computed. In FY 2003, for example, per capita expenditures were reported to be lowest in Wyoming ($1,583), Puerto Rico ($2,691), Washington ($3,150), New Mexico ($3,222), Kansas ($3,409), Nebraska ($3,559), Minnesota ($4,136), North Dakota ($4,488), the U.S. Virgin Islands ($4,488), and Missouri ($4,764). These States and territories also vary widely in enrollment, total funds expended, sources of ADAP funding (including total Title I funds allocated), eligibility requirements, formulary, and payment for insurance-related expenses. Some of these States and territories have been able to reduce their per capita expenditures because they purchase insurance premiums for medications and other health care benefits. Other States and territories with low per capita expenditures have relatively small ADAP budgets that are dedicated primarily to medication purchasing.

Factors Contributing to Variation in ADAP Per Capita Expenditures

GWU explored the association between ADAP per capita expenditures and geographic location, State ADAP characteristics, the characteristics of Medicaid programs in the States in which the ADAPs operate, and the external State health care environment. Table 2 (see page 9) outlines those factors. A significant, negative association was found to exist between State per capita expenditures and the total co-payments and deductibles paid per year by State ADAPs. The higher the per capita State ADAP expenditures, the lower the total co-payments and deductibles paid per year by State ADAPs. Conversely, the higher the per capita State ADAP expenditures per year, the lower the percentage of the State population with individual commercial insurance.

Results of statistical analyses underscore the importance of a close collaborative relationship between State ADAPs and the Medicaid program to ensure that people eligible for Medicaid benefits rapidly enroll in that program. State ADAP per capita expenditures were lower among programs that had a Medicaid online interface with which Medicaid enrollment could be verified ($6,617) than among State ADAPs that did not have this interface ($7,862). Note that several large and medium-sized State ADAPs still do not have an electronic interface in place because of technical issues or lack of interagency agreement about the feasibility or legality of such an interface. State ADAP per capita expenditures were higher among programs with dual applications for Medicaid and ADAP enrollment ($8,281) than among State ADAPs without dual applications ($6,649). State ADAP per capita expenditures were higher among programs requiring proof of Medicaid application denial ($7,758) than among State ADAPs without this requirement ($6,411). A nonsignificant weak association was found between State ADAP per capita expenditures and joint ADAP and Medicaid administration at the State level; States with joint administration had lower average per capita expenditures ($5,828) than State ADAPs without joint administration ($7,188).

Statistical analysis also pointed to the importance of State ADAPs retroactively billing Medicaid for ADAP clients that are subsequently awarded retroactive Medicaid pharmaceutical coverage. State ADAP per capita expenditures were lower among programs with retroactive Medicaid billing ($6,041) than among State ADAPs that did not have this billing process ($7,808).

GWU staff explored the relationship between per capita State ADAP expenditures and receipt of Title I funds. In FY 2003, 12 States or territories (22 percent of State ADAPs) reported receiving Title I funds. In FY 2003, Title I funds received by State ADAPs ranged from $57,425 to $18 million and averaged $426,924. The number of State ADAPs receiving Title I funds increased between FY 2002 and FY 2003, with 12 State ADAPs receiving Title I funds in FY 2003, compared with 8 State ADAPs in FY 2002. The average per capita amount was $8,359 for States receiving Title I funds, compared with $7,125 for States that did not receive Title I funds. This difference was not statistically significant, however.

Discussion

While conducting this study, State ADAPs continued to make rapid changes in their program designs and policies that were not reflected in the data available to GWU staff, including the following:

• Setting further limits on income and assets
• Adjusting the frequency of recertification (monthly, quarterly, biannually, or annually, as the formulary was changed)
• Requiring Medicaid application or proof of denial
• Establishing online interface with Medicaid to reduce dual enrollment
• Increasing assessment of case management eligibility
• Denying benefits to veterans
• Establishing waiting lists and capped enrollment
• Prioritizing applicants on the basis of clinical factors (pregnancy, CD4, or viral load)
• Reducing the formulary to exclude drugs not directly related to HIV treatment
• Establishing a monthly ceiling on the number of drugs paid for by the State ADAP
• Setting limits on protease inhibitors (PIs) or opportunistic infection medications
• Attempting to expand rebate agreements and reduced prices
• Establishing co-payments for ADAP clients
• Increasing the purchase of insurance premiums
• Reorganizing distribution systems to reduce dispensing costs
• Implementing retrospective Medicaid billing.

TOP