UCSF

Objectives

To examine how dual coverage for nonelderly, low-income veterans by Veterans Affairs (VA) and Medicaid affects their demand for care.

Data sources

Veterans Affairs utilization data and Medicaid Analytic Extract Files.

Study design

A retrospective, longitudinal study of VA users prior to and following enrollment in Medicaid 2006-2010.

Data collection/extraction methods

Veterans Affairs reliance, or proportion of care provided by VA, was estimated with beta-binomial models, adjusting for patient and state Medicaid program factors.

Principal findings

In a cohort of 19,890 nonelderly veterans, VA utilization levels were similar before and after enrolling in Medicaid. VA outpatient reliance was 0.65, and VA inpatient reliance was 0.53 after Medicaid enrollment. Factors significantly associated with greater VA reliance included sociodemographic factors, having a service-connected disability, comorbidity, and higher state Medicaid reimbursement. Factors significantly associated with less VA reliance included months enrolled in Medicaid, managed care enrollment, Medicaid eligibility type, longer drive time to VA care, greater Medicaid eligibility generosity, and better Medicaid quality.

Conclusion

Veterans Affairs utilization following new Medicaid enrollment remained relatively unchanged, and the VA continued to provide the large majority of care for dually enrolled veterans. There was variation among patients as Medicaid eligibility and other program factors influenced their use of Medicaid services.