Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website
2015 Conference Logo



2015 HSR&D/QUERI National Conference Abstract


3144 — Veterans' Pharmacy and Healthcare Utilization following Implementation of the Medicare Part D Pharmacy Benefit

Stroupe KT, Hines VA Hospital; Bailey L, Hines VA Hospital; Gellad WF, VA Pittsburgh Healthcare System; Suda K, Hines VA Hospital; Huo Z, Hines VA Hospital; Martinez R, Hines VA Hospital; Burk M, VA Pharmacy Benefit Management Services; Cunningham F, VA Pharmacy Benefit Management Services; Smith BM, Hines VA Hospital;

Objectives:
Because > 40% of VA users are Medicare enrolled, a substantial portion of VA users are eligible for Medicare Part D drug benefits, which became available January 2006. We examined the association of Part D enrollment with changes in VA pharmacy use and VA and Medicare-reimbursed healthcare use.

Methods:
We included all women and a random 10% sample of men enrolled in VA who were > 65 years old by January 2004 and alive through December 2007. A difference-in-differences approach was used to examine changes in utilization between 2005 (before) and 2007 (after Part D was implemented) by Part D-enrolled Veterans compared to non-enrollees. Pharmacy use was measured as the number of 30-day equivalent medication supplies. We implemented this using propensity score methods within GEE models.

Results:
Of 181,997 Veterans, 49,882(21.5%) were enrolled in Part D during 2007. VA 30-day supplies decreased from 26.2 to 23.4 among enrollees but increased from 36.6 to 37.4 among non-enrollees (weighted difference-in-differences: -4.0, p < 0.001), indicating a 15% (4/26.2) decrease. Enrollees with copayments for some or all VA medications had -4.6 (99%CI:-5.2,-4.0) and -4.0 (99%CI:-4.4,-3.5) reductions in 30-day supplies compared to non-enrollees, representing 14% and 23% decreases. Non-overlapping 99%CIs indicated these reductions were significantly different than for Veterans without VA copayments (0.19;99%CI:-2.6,3.0). For enrollees with no, low, moderate, or high VA pharmacy use in 2005, there were reductions of -0.5 (99%CI:-0.8,-0.3), -3.5 (99%CI:-4.2,-2.8), -7.8 (99%CI:-8.6,-7.0), and -8.0 (99%CI:-10.1,-6.3) 30-day supplies compared to non-enrollees. Compared to non-enrollees, enrollees dually using VA/Medicare outpatient services in 2005 had a greater decrease in VA pharmacy use (-6.4, 99%CI:-7.0,-5.7) than VA-only (-0.9, 99%CI:-1.2,-0.6) or Medicare-only (-3.6, 99%CI:-5.2,-2.0) outpatient users. The number of VA outpatient visits decreased, while the number of Medicare-reimbursed outpatient visits increased for enrollees compared to non-enrollees.

Implications:
Veterans' enrollment in Medicare Part D was associated with reductions in medications obtained from VA pharmacies and changes in VA and non-VA healthcare use.

Impacts:
As healthcare reform efforts proceed with implementation of the ACA and the Choice Act, it is important for clinicians and policymakers to consider that enhanced access to non-VA services may impact patterns of healthcare use across systems of care.