Study 1
RYAN WHITE CARE ACT REAUTHORIZATION 2005: TITLE I AND TITLE II HEALTH SERVICES EXPENDITURES PATTERNS
RUTH FINKELSTEIN
Purpose of the Study
To investigate whether the allocation of Ryan White Comprehensive AIDS Resources Emergency (CARE) Act funds for health care services by States and eligible metropolitan areas has changed through time and whether such allocations are responsive to differences in the HIV health care environments. The study hypothesized that Title I and Title II proportional expenditures on health care services would increase over time and that CARE Act Title I and Title II jurisdictions with better resourced HIV health care environments would spend proportionately less CARE Act money on health care services because those services would be covered by other payers. It also hypothesized that jurisdictions with poorly resourced HIV health care environments would spend more CARE Act money on health care services to make up for such deficiencies.
Background
One central tenet of the Ryan White CARE Act is that flexibility and decision making at the local level are essential to respond to variations in the needs, capacity, and resources in different jurisdictions. This local control and flexibility have been driven by the recognition that the existing services and resources available to care for people living with HIV/AIDS (PLWHA) are different in different places. Moreover, historically, much emphasis has been placed on the differing characteristics of HIV epidemics in different locations, which have been driven by different modes of transmission and effects on different populations. As medical treatments for HIV have improved, however, and as these “different” HIV epidemics have increasingly converged, it has become possible to define a continuum of care for PLWHA throughout the country. Local planning in Titles I and II (and grants through Titles III and IV) can be conceptualized to identify specific, unmet needs of PLWHA in the jurisdiction and existing resources and capacities to meet those needs and to prioritize CARE Act funds to fill identified gaps. Logically, therefore, a structured relationship should exist between the existing local HIV health care environment and the services funded through the CARE Act.
Allocation of Title I and II resources to fund health care services represents a critical opportunity to investigate the appropriateness of local planning decisions for a number of reasons. First, since the enactment of the CARE Act, and particularly since the advent of highly active antiretroviral therapy in 1996, both the effectiveness and the cost of HIV primary care have increased because consistent participation in HIV primary care enhances longevity and decreases morbidity and mortality.1,2,3 Local health care environments vary from State to State and even within States. For HIV care and services, this variation is driven principally by enormous variation in State Medicaid programs (including covered populations, covered services, and reimbursement structures and levels).4 Medicaid is estimated to cover 44 percent of people living with HIV and 55 percent of people living with AIDS (PLWA) nationwide.4 Second, local HIV health care environments vary by the availability of publicly funded HIV health care (based on the existence of public hospitals and community health centers) and existence and the quality of HIV care capacity as well as the access of PLWHA to private insurance. Thus, local jurisdictions differ in their capacity to provide HIV health care services and the quality, accessibility, and affordability of such services. This variation makes it possible to examine whether differences in local health care environments are related to different decisions about how to allocate Title I and Title II funds.
Methodology
To explore these hypotheses, this study looked at trends in Title I and Title II health care spending over a 3-year period, analyzed the relationship between health care spending and variables in local health care environments, constructed a typology of health care environments and spending patterns, and conducted case studies to better understand the relationships between the health care environment and CARE Act spending. The dependent variable (health care services spending) was defined broadly in accordance with Health Resources and Services Administration (HRSA) guidance. Because the focus was on discretionary spending, the trend analysis and the health care environment analysis excluded funds earmarked for specific services, such as the AIDS Drug Assistance Program (ADAP) earmark, and included all discretionary spending from the base award, carryover, Minority AIDS Initiative, and emerging communities awards in the dependent variable. The ADAP earmark was, of course, included in the independent variables describing the health care environment. Independent variables describing the HIV health care financing environment were based on secondary sources that could provide uniform data across jurisdictions. The study examined characteristics of Medicaid, other CARE Act titles, the adequacy of the ADAP program, and other major Federal programs providing PLWHA health care services, including the Department of Veterans Affairs, National Institutes of Health, Indian Health Service, and Medicare, and estimates of the proportion of AIDS-related services covered by private insurance.
The trend analyses focused on Title I and II expenditure data submitted to HRSA for fiscal years (FYs) 2000 through 2002, the most recent years for which data were available. Data were analyzed for trends in expenditures related to broad service categories—health care services, support services, and nonservice categories—and for the proportion of total expenditures related to each category. The study then examined the relationship between health care spending in FY 2002 and the health care financing environment. Results from this analysis were used to create a typology of the HIV health care environments so that CARE Act jurisdictions could be characterized across a uniform set of variables. Finally, case studies of 12 jurisdictions were conducted to illuminate how planning and resource allocation by grantees respond to the HIV health care environment and to changes in that environment. In addition, the case studies provide information that helps interpret other findings about trends in health care spending and the relationship between the health care environment and spending by Title I and Title II grantees.
Limitations
The study findings are subject to several limitations. Trend analysis data were incomplete and inconsistent in several cases. Moreover, it is likely that different jurisdictions attributed expenditures to service categories differently. Indeed, a related limitation of the trend analysis is that it relies on only three data points. The health care environment analysis also was subject to limitations: because uniform secondary data sources were lacking, it was impossible to include all key features of the HIV health care environment. Even when uniform data were available, the data sets were not always complete or consistently reported. The typology was created by transforming a continuous variable (the health care environment resource score) into a categorical variable (high, medium, low) in order to sort the jurisdictions meaningfully. Doing so, however, created the possibility that the “bottom” of one group is more closely related to the “top” of another group than to the mean of the group in which it was classified. Finally, the health care environment score and the typologies are based on cross-sectional data for FY 2002. As the case studies confirmed, this snapshot in time fails to capture recent changes in the environment or in how planning decisions might have responded to such changes.
Despite its limitations, the study supports some general conclusions about trends in spending and how key features of the HIV health care environment relate to proportional spending of CARE Act funds on health care services.
Major Findings
Health Care Spending Trends
No significant changes were observed in the proportional spending of discretionary Title I and Title II funds on health care services from FY 2000 to FY 2002. Although the total money spent on health care increased, proportional health care expenditures did not increase as expected. This finding is consistent with and lengthens the trend line developed by Young et al. in their analysis of FY 1996 through FY 2000 expenditure data.5 Significant variation was found among jurisdictions in proportional health care services spending as well as, in some jurisdictions, temporal changes. When the ADAP earmark was included in the analysis, proportional health care spending did increase each year. However, this increase is driven by the earmark and Federal regulation of the CARE Act, not State and local decision making. Because the study focused on local and State discretionary spending, the ADAP earmark was excluded from subsequent analyses (except as a component of the health care environment). Although no trend emerged in health care services spending, systematic trends were found in specific service categories, particularly decreased spending on services associated with care for the acutely ill (e.g., home health care, hospice services). This finding, also consistent with those of Young et al., likely reflects the decline in prevalence of PLWHA with end-stage disease. Despite this minor trend, the big picture was revealed to be one of relative stasis in proportional spending.
Relationship Between the HIV Health Care Environment and Proportional Health Care Expenditures
The analysis sought to identify significant relationships between individual features in the health care financing environment (Medicaid, other insurance, other CARE Act titles, the adequacy of the ADAP program, and other major Federal programs providing health care to PLWHA) and health care services expenditures in FY 2002. Basic demographic variables (percentage of population living in poverty and number of PLWA) were included to roughly characterize the HIV epidemic in each jurisdiction. All of these variables were then used to create a composite typology that characterized jurisdictions as having a high, medium, or low HIV health care financing environment. The study also explored the relationship between these types and CARE Act health care service expenditures.
Analysis of the relationship between individual variables in the health care environment and CARE Act health care expenditures yielded few relationships. For Title I, no significant relationships were detected between any characteristics of the health care environment and health care expenditures. For Title II, relationships were found in the expected direction for two characteristics. States that had higher income thresholds for people to qualify for their Medicaid medically needy program and States that had no medically needy program spent more Title II money on health care services than did States with a lower threshold for eligibility for medically needy programs. In addition, States that had a higher number of special Medicaid programs for PLWHA had lower proportional Title II spending on health care services than did States with few or no special programs.
| Health Care Environment Score |
Title II | Title I | ||
|---|---|---|---|---|
| Low | Oregon | 14.2 | Portland, OR | 24.4 |
| Nebraska | 16.3 | Las Vegas, NV | 36.7 | |
| Hawaii | 29.0 | St. Louis, MO | 38.0 | |
| Arkansas | 31.3 | Kansas City, MO | 38.6 | |
| Vermont | 37.7 | West Palm Beach, FL | 39.9 | |
| Alaska | 38.7 | New Orleans, LA | 40.1 | |
| Kansas | 39.0 | Fort Worth–Arlington, TX | 42.2 | |
| Texas | 40.6 | Norfolk, VA | 54.4 | |
| Colorado | 42.3 | Austin, TX | 54.5 | |
| South Carolina | 46.5 | San Antonio, TX | 59.3 | |
| Kentucky | 46.7 | Denver, CO | 68.7 | |
| Montana | 49.7 | |||
| Idaho | 54.0 | |||
| Alabama | 54.4 | |||
| New Mexico | 71.4 | |||
| Wyoming | 78.5 | |||
| Indiana | 85.0 | |||
| Nevada | 87.5 | |||
| Medium | Maine | 00.1 | Minneapolis–St. Paul, MN | 8.6 |
| Washington | 19.5 | Vineland–Millville–Bridgeton, NJ | 24.6 | |
| Missouri | 30.1 | Sacramento, CA | 34.2 | |
| Wisconsin | 35.0 | Seattle, WA | 35.2 | |
| Iowa | 38.0 | Hartford, CT | 36.4 | |
| North Carolina | 39.9 | Jersey City, NJ | 38.1 | |
| Rhode Island | 44.2 | Orange Co., CA | 39.5 | |
| Utah | 47.2 | Phoenix, AZ | 40.3 | |
| North Dakota | 47.4 | Riverside–San Bernardino, CA | 48.2 | |
| Delaware | 51.9 | Middlesex–Somerset–Hunterdon, NJ | 51.0 | |
| Louisiana | 55.5 | Jacksonville, FL | 51.9 | |
| Virginia | 55.5 | San Jose, CA | 53.1 | |
| Arizona | 63.7 | Orlando, FL | 56.0 | |
| New Hampshire | 67.6 | Houston, TX | 57.8 | |
| Tennessee | 72.1 | Cleveland–Lorain–Elvira, OH | 58.7 | |
| Oklahoma | 76.9 | Dallas, TX | 62.7 | |
| Mississippi | 88.7 | Fort Lauderdale, FL | 63.3 | |
| Tampa–St. Petersburg, FL | 68.8 | |||
| High | Massachusetts | 16.5 | Oakland, CA | 23.4 |
| Pennsylvania | 16.7 | Boston, MA | 24.5 | |
| Connecticut | 26.9 | San Diego, CA | 29.0 | |
| District of Columbia | 37.6 | Detroit, MI | 31.5 | |
| Florida | 50.7 | Philadelphia, PA | 36.0 | |
| Illinois | 51.2 | Newark, NJ | 41.1 | |
| New York | 52.4 | New Haven, CT | 42.9 | |
| Minnesota | 58.6 | Baltimore, MD | 44.5 | |
| Michigan | 62.3 | Washington, DC | 46.1 | |
| Ohio | 66.3 | Santa Rosa–Petaluma, CA | 47.7 | |
| Maryland | 69.3 | Chicago, IL | 52.3 | |
| California | 73.0 | Nassau–Suffolk, NY | 52.3 | |
| New Jersey | 74.2 | San Francisco, CA | 52.7 | |
| Georgia | 80.3 | New York, NY | 54.6 | |
| Bergen–Passaic, NJ 56.7 | ||||
| Duchess County, NY 58.3 | ||||
| Miami, FL 62.9 | ||||
| Los Angeles, CA 67.3 | ||||
| Atlanta, GA 74.4 | ||||
As Table 1 illustrates, when jurisdictions are classified into broad types of health care environments, a full distribution of CARE Act health care spending may be found within each type. The typologies further demonstrate the absence of a structured or predictable relationship between local health care financing environments and the proportion of Title I and Title II funds expended on health care services. The case studies did, however, reveal instances of spending that were responsive to the health care environment but were not detected by the statistical analysis. For example, several jurisdictions used carryover funds to address shortfalls in other funding, such as ADAP or coverage for a specific test disallowed by Medicaid.
Table 1 shows clearly that in each health care environment—high, medium, and low—discretionary expenditures on health care services vary dramatically, from lows of under 20 percent of the award to highs of more than 80 percent in Title II programs and from less than 25 percent to more than 65 percent in Title I programs. Table 2 confirms that no trend or pattern toward a relationship exists between proportional health care spending and the adequacy of the health care financing environment. The slight nonsignificant trend suggested goes in the wrong direction for both Title I and II—that is, the more resources in the health care environment, the greater the proportional spending on health care services from Titles I and II.
| Title I Health Care Typology Score |
|||||
|---|---|---|---|---|---|
| Low | Medium | High | F | p | |
| N | 11 | 18 | 19 | 0.082 | ns |
| Mean | 45.2 | 46.0 | 47.6 | ||
| Standard Deviation | 12.6 | 15.2 | 14.1 | ||
| Title II | |||||
| N | 18 | 17 | 14 | 0.324 | ns |
| Mean | 47.9 | 49.0 | 53.7 | ||
| Standard Deviation | 21.2 | 21.8 | 19.9 | ||
ns=not significant. |
|||||
Recommendations
Policy Implications and Options
Resource allocations for Title I and, to a lesser extent, Title II are relatively impervious to local health care financing conditions.
Possible Policy Options
- Reduce the local planning and resource allocation functions in Title I or in Titles I and II of the CARE Act. Such action would require legislative change.
- Continue to support local planning but establish more comprehensive guidance about how it should be accomplished.
- Continue local planning but legislatively require a hierarchy of allocations to ensure the prioritization of health care services.
Comprehensive planning at the State and local level is difficult and complex.
Possible Policy Options
- Reexamine what information and knowledge are expected at what levels of government and planning.
- Develop a comprehensive, Web-based resource guide for all facets of comprehensive planning, including links to local data and resources.
- Reduce the administrative burden of planning by recognizing that new systems of planning and coordination should replace or integrate existing ones.
- Work with grantees to improve collaboration between Title I and Title II within each State.
Discretionary planning focuses primarily on increases from the previous award rather than on the whole award.
Possible Policy Options
- Eliminate all restrictions, earmarks, and set-asides within Title I and Title II. Such action would require legislative change.
- Require zero-based planning at regular intervals so that grantees plan for and reallocate the entire portfolio.
References
1 Marconi KM, Jacobsen JM. Trends in the use of primary health care and case management among HIV-positive individuals. [Abstract]. Assoc Health Services Res. 1999;16:160.
2 Mocroft A, Ledergerber B, Katlama C, et al. Decline in the AIDS and death rates in the EuroSIDA study: An observational study. Lancet. 2003;362(9377):22–29.
3 Bartlett JG, deMasi R, Quinn J, Moxham C, Rousseau F. Overview of the effectiveness of triple combination therapy in antiretroviral-naive HIV-1 infected adults. AIDS. 2001;15(11):1369–1377.
4 Institute of Medicine. Public Financing and Delivery of HIV/AIDS Care: Securing the Legacy of Ryan White. Washington, DC: National Academies Press; 2004.
5 Young S, Conviser R, Marconi KM, Wieland M. Trends and responsiveness in national resource allocation for needed HIV services: A five-year (1996–2000) analysis. J Health Soc Policy. 2003;17(4):1–14.
