IIR 07-165
Impact of Medicare Drug Benefit on VA Drug Use, Healthcare Use and Cost
Kevin T. Stroupe, PhD MA BS Edward Hines Jr. VA Hospital, Hines, IL Hines, IL Funding Period: January 2011 - June 2013 |
BACKGROUND/RATIONALE:
Over 40% of Veterans eligible for VA healthcare are enrolled in Medicare. Beginning January 2006, drug coverage was available to all Medicare enrollees through Medicare Part D. Individuals with Medicare coverage, including Veterans receiving care at VA, could begin enrolling in Part D from 11/2005-5/2006. For those who enrolled in Part D, coverage began January 2006 or the first day of the month following enrollment. Subsequent enrollment periods were November through December of each year with coverage beginning the following January. Many dual (VA and Medicare) users may be receiving care at the VA primarily to obtain the VA pharmacy benefit, which requires a prescription from a VA physician. Thus, the availability of Part D could lead some Veterans to alter their patterns of use of VA and Medicare-covered services. Consequently, the extent of Veterans' enrollment in Part D and the overall impact of Part D on medication acquisition from VA and healthcare use and costs in VA and Medicare were unknown. OBJECTIVE(S): The primary objectives were to determine 1) the extent of Veterans' enrollment in Part D and 2) the impact of Part D on use of VA and Medicare-covered services. METHODS: We identified Veterans who met Medicare's age eligibility in 2004 (i.e., Veterans who were 65 years of age or older on January 1, 2004) in the VA Vital Status File. We took a random 10% sample of male and all female Veterans from this file who were 65 years of age or older as of January 1, 2004, and were presumed alive as of December 31, 2006. Based on these eligibility criteria, we identified 327,100 Veterans: 270,445 male and 56,655 female Veterans. For the first objective, we examined the proportion of our cohort who were enrolled from 2006 through 2009, and we compared characteristics of those who enrolled vs. did not enroll. For the second objective, we compared medication acquisition from VA, VA and Medicare service use in 2005 (before Part D was implemented) and 2007 (after Part D was implemented) between Veterans who were or were not enrolled in Part D. We excluded 2006 because this start-up period may not be representative of the on-going program. We identified the association of the Part D program with these outcomes using a difference-in-differences approach by comparing the difference in outcomes between 2007 and 2005 for Veterans who did and did not enroll in Part D. We implemented the difference-in-differences approach by first matching patients who were or were not enrolled in Part D in 2007 using propensity scores. Then, we estimated the association of Part D with our outcomes using generalized estimating equations (GEE). We identified the difference-in-differences using the interaction term of enrollment and time in the GEE model. Because of potential gender differences in medication and healthcare utilization patterns, we examined men and women separately. FINDINGS/RESULTS: There were 320,019 Veterans with complete data who were included in the enrollment analyses. Veterans averaged 77.5 (6.1 SD) years old in 2006. Overall, 31.9% enrolled in Part D in 2006, and 40.9% were enrolled some time between 2006 and 2009. Veterans who were enrolled in Part D in 2006 had significantly higher odds than Veterans not enrolled of being male (OR=1.11, p<0.001), Hispanic (OR=1.34, p<0.001), and having higher co-morbidity scores (OR=1.20, p<0.001). Older Veterans also had higher odds of enrollment; Veterans between the ages of 71-75 had 1.17 higher odds of enrolling in 2006 than younger Veterans. Among Veterans who enrolled in 2006, 7.7% quit in 2007, and an additional 3.6% (3.6%) quit in 2008 (2009). Characteristics significantly associated with quitting in 2007 included having higher numbers of VA prescriptions in 2006 (OR=1.01, p<0.001), being African-American (OR=1.23, p<0.001), and having a lower co-morbidity score (OR=0.81, p<0.000). There were 236,764 men (50,141 women) with complete data who were alive throughout 2007. We performed one-to-one nearest neighbor propensity matching, resulting in 75,120 men (15,766 women) enrolled in Part D in 2007 and a similar number who were not enrolled. We examined the number of 30-day equivalent supplies of medications that Veterans received from VA pharmacies. Consistent with our hypothesis of decreased VA pharmacy use for enrolled Veterans, difference-in-differences testing revealed that the number of 30-day drug supplies increased from 31 in 2005 to 35 in 2007 for male Veterans not enrolled but decreased from 31 in 2005 to 29 in 2007 for enrolled Veterans (p<0.001 both). GEE analysis indicated a 17% reduction in medication acquisition from VA pharmacies for male Veterans enrolled in Part D (incidence rate ratio [IRR]=0.83, p<0.001) relative to those not enrolled. Similarly, there was a 19% reduction in medication acquisition from VA pharmacies for female Veterans enrolled in Part D (IRR=0.81, p<0.001). Consistent with our hypothesis that Veterans who enrolled in Part D decreased utilization of VA services, we found that VA primary care visits decreased by 14% (IRR=0.865, p<0.001) for men and 15% for women (IRR=0.852, p<0.001) who enrolled relative to those not enrolled. VA inpatient length of stay also decreased by 6% (IRR=0.936, p<0.001) for men and 22% for women (IRR=0.784, p<0.001). However, the findings about Part D enrollment and use of Medicare-covered services were not consistent. For both men and women, Medicare-covered primary care decreased for patients enrolled in Part D relative to those not enrolled but the length of stay for Medicare-covered inpatient admissions increased. IMPACT: Between 2006-2009, 40.9% of patients meeting Medicare's age eligibility were enrolled in Part D. While Part D may increase options for obtaining medications among Veterans using VA healthcare, non-VA medication coverage might also complicate coordination among multiple prescribers. Through a previous survey of Veterans, we found that a substantial portion of Veterans were not discussing non-VA medications with VA physicians. Consequently, our current results highlight the importance of medication reconciliation efforts to account for non-VA pharmacy use, especially among Medicare-eligible VA users. External Links for this ProjectDimensions for VADimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.Learn more about Dimensions for VA. VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project PUBLICATIONS:Journal Articles
DRA:
Health Systems
DRE: Treatment - Observational, Treatment - Comparative Effectiveness Keywords: Policy - VA or other MeSH Terms: none |