The Impact of Medi-Cal Restructuring on Care of Beneficiaries with HIV/AIDS

William Cunningham, UC-Los Angeles
HIV/AIDS Policy & Health Care Financing

Background: Medicaid is the principal payer of care for Americans living with advanced HIV-infection. Individual state Medicaid programs have promoted managed care plans (MCPs) over traditional Fee-For-Service (FFS) for controlling healthcare costs for individuals with chronic disease, such as persons living with HIV. California has the second largest cohort of HIV-infected Medicaid enrollees in the country and has introduced MCPs on a county-by-county basis. We examine the impact of MCP enrollment on mortality, hospitalization, changes in enrollment and cost among Medicaid beneficiaries with AIDS.

Methods: This study uses a retrospective longitudinal cohort study of 12, 078 Medicaid beneficiaries with AIDS in urban counties of California, enrolled in January 1, 1999 and followed through December 31, 2003. We are analyzing Medi-Cal enrollment data linked to Medi-Cal claims, state hospital discharge abstracts, the state death certificate registry and the state AIDS Registry. The impact of MCP enrollment on mortality, hospitalization and changes in enrollment is estimated using multivariate regression models.

Results: In preliminary results, we identify 12, 078 individuals, with 14. 5% enrolled in MCPs. Characteristics of enrollees were: 83% male, age 41. 5 years old (mean), 4. 7 years since AIDS diagnosis (mean),46% non-LatinoWhite, 28% non-Latino Black, and 24% Latino. One third of the enrollees were hospitalized in 1998. During the study period, 63. 3% of enrollees were hospitalized and 23. 3% died. Rates were similar among MCP and FFS enrollees. However, 24. 3% of MCP enrollees changed to FFS by the end of the period, while only 5. 6% of FFS enrollees changed to a MCP. Multivariate regression models found no significant relationship betweenMCPenrollmentand mortality, but MCP enrollment did appear to be associated with lower rates of hospitalization. MCP enrollment was associated with a greater odds of changing plans (Odds Ratio: 6. 81, 95% Confidence Interval: 5. 71 to 8. 13).

Conclusions: Similar death rates with lower utilization among managed care recipients suggest that MCP enrollment may reduce hospitalization rates. Unmeasured severity may bias these estimates, suggesting benefit where there is none. Greater likelihood of disenrollment from MCPs is suggestive of low patient satisfaction or difficulties with access to care. Before wholesale changes are made in the delivery of care, policy makers should address whether non-clinical measures such as cost, satisfaction and access to care differ substantially between MCP versus FFS Medicaid for HIV-infected and other chronically ill beneficiaries. Future work will focus on revising estimates of the impact of MCP on outcome, comparing the cost of care for current MCP and FFS enrollees and examining physician readiness where mandatory MCP enrollment would be instituted.