Understanding the variation in costs among HIV primary care providers

Abstract

The paper uses a hybrid cost model to identify the determinants of cost variation among programs that offer early intervention services to people living with HIV and AIDS in the US. The model combines the effects of input price and output volume measures from traditional economic cost functions with institutional factors based on program and patient characteristics on the cost of providing primary medical care and support services to people living with HIV and AIDS. The impact of economic factors conforms to conventional theory and reveals the potential for cost savings through greater economies of scale and substitutability of low cost for high cost labor inputs. Similarly, programs that use staff more efficiently and share an affiliation with other organizations exhibit lower costs than more labor intensive and non-affiliated providers. However, patient characteristics are equally important determinants of program spending. Minority patients use services less frequently and generate fewer costs, while patients facing fewer barriers to care, such as those with Medicaid coverage, access services more frequently and incur higher costs. Uninsured patients also generate higher costs, but the higher costs associated with this subgroup more likely stem from a lack of continuity in care and, thus, poorer health status and greater healthcare needs when treatment is sought. Injection drug users require less expensive services, but access services more frequently than other risk groups, while patients with an AIDS diagnosis and those who are co-infected with hepatitis C require more program resources. By separately estimating the economic and institutional determinants of program costs, the study highlights the relative importance of factors that are amendable to internal cost control efforts versus those that reflect the resource needs of local communities.

Keywords:

• HIV/AIDS
• primary care
• healthcare costs
• healthcare financing

Acknowledgements

This research in this article was supported by the HIV/AIDS Bureau in the Health Resources and Services Administration (HRSA) under Contract Number 250-01-008. The statements expressed in this article are those of the authors and do not necessarily reflect the views or policies of HRSA. The authors would like to thank Dr. Richard Conviser for his substantive contributions throughout the study.

Notes

1. In 2006, the Ryan White Comprehensive AIDS Resources Emergency (CARE) Act of 1990 [P.L. 101–381] and the 1996 amendments [P.L. 104–146] and 2000 [P.L. 106–345] were reauthorized under the Ryan White HIV/AIDS Treatment Modernization Act of 2006 [P.L. 109–415]. At the same time, Title III of the original Act (under which the EIS program is implemented) was renamed Part C. Because the study is based on data from the period before the 2006 reauthorization, we refer to the Act and title program under their earlier names.

2. The CADRs have recently been renamed and are now referred to as the Ryan White HIV/AIDS Data Report.

3. The outlier exclusion criterion results in a loss of nine providers (out of a total of 264) from the 2002 group and eight providers (out of a total of 337) from the 2003 group The models were run on the full sample and the results were generally consistent with those presented in this study.

4. This HIV caseload exclusion criterion results in a loss of 23 providers from the 2002 group and 34 providers from the 2003 group.

5. A normal probability plot revealed that program expenditures and input costs were non-normally distributed. The normality assumption was rejected using a Shapiro-Wilk test.

6. Most clients with Medicare coverage will also be enrolled in Medicaid as a secondary payer.