BACKGROUND/RATIONALE:
The IOM recommends that the federal government try to understand how well its health care programs fit together for patients who are dual- or triple-enrolled. VA researchers recognize the importance of dual enrollment in VA and Medicare but have not examined VA-Medicaid dual-enrollment (6.5% of all VA patients).

OBJECTIVE(S):
Objective 1: Relate the use of VA, Medicaid and Medicare services nationally for all veterans dually enrolled in VA and Medicaid (VA+Medicaid enrollees) to the patients' needs.
Objective 2: Identify Medicaid program factors leading to greater VA+Medicaid enrollment.
Objective 3: Compare risk-adjusted outcomes for VA+Medicaid enrollees and comparable VA-only patients
Objective 4: Describe and compare the patterns of pharmacy care for dual users and veterans who use only the VA or Medicaid for their prescription drugs.

METHODS:
This study used Medicaid national eligibility and utilization files for 1999 and 2000, VA national utilization, outpatient PBM pharmacy data, Medicare inpatient MEDPAR, SNF, outpatient, and other standard analytic files, and the BIRLS files. We selected all VA+Medicaid enrollees from 1999 and 2000 using national MSIS Eligible files and a master file of VA patients created from multiple years of VA utilization data to identify VA patients also enrolled in Medicaid. Matching Medicare Denominator files gave the triply enrolled subpopulation.
Estimates of health expenditures used HERC national averages, VA disbursed amounts for contract care delivered in the study years, and DSS national aggregate outpatient pharmacy costs. Value of Medicare and Medicaid utilization was the total amount paid, including patient liability.
We compared total numbers of inpatient days and outpatient visits and their dollar values for the dually (and triply) enrolled to other VA patients with similar diagnoses or types of care (e.g., nursing home stays). We created a typology of state Medicaid characteristics relevant to the VA population and documented the VA programs available on a state level, using these in analyses of Medicaid enrollment among VA patients. We compared risk-adjusted mortality and preventable hospitalizations for dual enrollees to other VA patients with the same diagnoses to test for poorer outcomes that might indicate poor coordination of care.

FINDINGS/RESULTS:
About 305,000 VA patients were enrolled in Medicaid programs nationwide in either 1999 or 2000. About a quarter of these patients were in nursing homes; a significant proportion had diagnosed mental health disorders. A disproportionate number, 15% of the total, qualified for VA as "Housebound, Aid and Assistance" as compared to the general VA patient population. Regression analysis suggests that patients using care through more than one system had an increased risk of preventable hospitalizations of about one third, raising it from about 2.5% to almost 4%.

IMPACT:
This study documents potential problems for vulnerable VA patients who access care in more than one health care system. Better clinical coordination is required to prevent or ameliorate problems with medications, preventive care and follow-up visits.

---

External Links for this Project

---

PUBLICATIONS:

Journal Articles

1. Shen Y, Hendricks A, Li D, Gardner J, Kazis L. VA-Medicare dual beneficiaries' enrollment in Medicare HMOs: access to VA, availability of HMOs, and favorable selection. Medical care research and review : MCRR. 2005 Aug 1; 62(4):479-95. [view]

DRA: Health Systems Science
DRE: none
Keywords: Access, Managed care, VA/non-VA comparisons
MeSH Terms: none