Medicare Part D spending was obtained from publicly reported data; Veterans Affairs (VA) spending data for these medications was obtained from VA national patient care databases; and VA prices for metoprolol succinate exceeded those in Medicare during 2016 because of VA contracting issues.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Venker B, Stephenson KB, Gellad WF. Assessment of Spending in Medicare Part D If Medication Prices From the Department of Veterans Affairs Were Used. JAMA Intern Med. 2019;179(3):431–433. doi:10.1001/jamainternmed.2018.5874
Several state and federal efforts to reduce prescription drug costs have proposed using the Department of Veterans Affairs (VA) as a model, given its ability to obtain deep discounts on medications through direct negotiation with pharmaceutical manufacturers and the use of a national formulary. Published studies using data from a decade ago estimated that Medicare Part D could save $14 billion to $22 billion annually if it paid prices similar to those paid by the VA,1-3 and a recent congressional report4 for 20 brand-name drugs estimated potential annual savings of $2 billion. However, none of these estimates used actual prices paid by the VA, which can be lower than published federal prices. Thus, we used the most recent national data available from Medicare and the VA to quantify the savings Medicare Part D would achieve if it paid the same prices for prescription drugs currently paid by the VA.
We analyzed publicly available Medicare Part D prescription data from the Centers for Medicare & Medicaid Services between January 1, 2011, and December 31, 2016.5 We obtained gross Medicare spending on each medication (defined based on the generic product name), quantity dispensed, and number of claims. We excluded drugs that had formulations other than capsules and tablets to accurately calculate per-unit costs and limited our analyses to the top 50 oral drugs dispensed based on Medicare spending. This project using deidentified data was exempted from institutional review board approval by the Veterans Affairs St Louis Health Care System.
To make Medicare and VA prescription spending comparable, we made 2 adjustments to Medicare spending. First, because publicly reported Medicare Part D data include patient payments and do not account for rebates, we discounted annual gross Medicare spending to account for these factors (eAppendix in the Supplement). Second, we subtracted a $2.50 dispensing fee for each Medicare claim because VA drug spending data do not include these costs.6 After these adjustments, medication spending was divided by the number of units dispensed to establish a Medicare per unit cost.
Using the database of all prescriptions dispensed by the VA from 2011 to 2016, we identified all matches for the Medicare medications and obtained comparable data on acquisition costs and quantity dispensed using the actual price paid by the VA. We then applied the per unit costs from the VA to Medicare to calculate the amount that Medicare would spend annually if it obtained VA prices for the same quantity of medications.
Annual net Medicare Part D spending on the top 50 oral drugs ranged from $26.3 billion in 2011 to $32.5 billion in 2016 (Table). In 2016, if Medicare Part D obtained VA prices, the cost of these medications would have been $18.0 billion, representing savings of $14.4 billion, or an estimated 44%. The projected magnitude of estimated annual savings from 2011 to 2015 was similar, ranging from 38% to 50%. The Figure shows actual Medicare Part D spending in 2016 for these 50 drugs and spending if Medicare obtained VA prices.
To our knowledge, this is the most up-to-date analysis that directly compares prices between Medicare Part D and the VA, accounting for estimated rebates and using actual VA prices as opposed to publicly reported federal supply schedule prices. Using these data, we calculated $14.4 billion in potential savings in 2016 for only 50 drugs.
There are important limitations to this study. First, we did not differentiate between dosages or long-acting formulations when calculating unit prices, but most medications were brand-only products and were typically priced similarly regardless of dose. Second, it is unlikely that Medicare could capture the entirety of these savings, as VA prices could increase once Medicare obtained access, and Medicare would likely require a national formulary to leverage these reductions in price. However, our analysis of potential savings is conservative by studying only 50 drugs, omitting many high-cost medications with injectable formulations (eg, insulin and adalimumab), and focusing on price only without accounting for changes in use that could also lead to savings.
Accepted for Publication: August 27, 2018.
Corresponding Author: Walid F. Gellad, MD, MPH, Division of General Internal Medicine, University of Pittsburgh, 200 Meyran Ave, Pittsburgh, PA 15213 (email@example.com).
Published Online: January 14, 2019. doi:10.1001/jamainternmed.2018.5874
Author Contributions: Dr Venker had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Venker, Stephenson.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Venker.
Administrative, technical, or material support: Venker, Gellad.
Supervision: Venker, Gellad.
Conflict of Interest Disclosures: None reported.
Disclaimer: This work represents the opinions of the authors alone and does not necessarily represent the views of the Department of Veterans Affairs or the United States Government.
Additional Contributions: Inma Hernandez, PharmD, PhD, University of Pittsburgh School of Pharmacy, and C. Bernie Good, MD, MPH, University of Pittsburgh School of Medicine, reviewed prior versions of this manuscript. They were not compensated for their assistance.