Estimates of the Number of Brand-name Drugs Affected by the Medicaid Rebate Cap in 2017

Abstract

This analysis of Medicaid reimbursement data examines the financial implications of removing the Medicaid rebate cap.

The Trump Administration has proposed uncapping Medicaid rebates to discourage drug price spikes. Drug manufacturers pay rebates to offset Medicaid pharmacy reimbursement, including a base rebate (23.1% of average manufacturer price [AMP]) and an inflation rebate to counterbalance price increases above inflation. The cap, which was implemented in 2010, limits the total rebate to the drug’s AMP, restricting manufacturer rebates for price increases more than 433% above inflation (ie, when the AMP equals 23.1% of AMP plus the difference between AMP and inflation-adjusted initial AMP). In this study, I estimated the number of brand-name drugs that had capped Medicaid payment rebates in 2017 and the total additional rebates without the cap.

Methods

This study was approved by The Pew Charitable Trusts. Public data were accessed and analyzed in July 2017.

I used Medicaid State Drug Utilization Data files (https://www.medicaid.gov/medicaid/prescription-drugs/state-drug-utilization-data/index.html) from 1991 through 2017 to estimate AMP (a proprietary average that must be estimated from market data). Because AMP is the average of manufacturer sales to retail community pharmacies (RCPs) and because Medicaid must reimburse pharmacies on the basis of actual acquisition cost,¹ Medicaid pharmacy reimbursement approximates AMP for brand-name drugs (estimated AMP). Medicaid reimbursement is within 4% of the National Average Drug Acquisition Cost,² which has been used to estimate AMP but is only available after December 11, 2013.³ Historically, states reimbursed above the acquisition cost¹; therefore, my initial AMP estimates may be overstated, which may underestimate the number of drugs that triggers the cap. The price history for the top 30 drugs by reduced rebate amount was manually reviewed using a commercial database for consistency between estimated AMPs and wholesale acquisition costs.

For brand-name drugs that are injected, implanted, inhaled, infused, or instilled (5i drugs), if more than 30% of sales are not to RCPs, manufacturers use a different AMP methodology that is not based on RCP sales.¹ I separately analyzed these drugs, only considering 5i drugs with a reported National Average Drug Acquisition Cost, a federal estimate of prices to RCPs (demonstrating RCP sales). Manual review removed drugs that cannot be or are rarely self-administered and likely use this alternate AMP methodology.

To estimate the number of drugs that triggered the cap, I identified the first full year of reimbursement data (ie, the second year of data) using a commercial database to link identifiers (National Drug Code) over time. Then, I calculated the inflation rebate for 2017 (the difference between the inflation-adjusted initial estimated AMP and the current estimated AMP) and identified all drugs whose total rebate exceeded the estimated AMP. I also calculated the amount above the cap (reduced rebate).

Results

Of the 1705 brand-name drugs assessed, 271 drugs (15.9%; 250 non-5i drugs and 21 5i drugs) at the National Drug Code level triggered the cap, reducing rebate payments to Medicaid by US$103 540 808 ($52 619 984 non-5i drugs and $50 920 914 5i drugs) (Table) in 2017. Seventy-five drugs (27.7%) had reduced rebates of greater than $100 000, and 85% of the reduced rebates were attributable to 25 drugs; diabetes treatments accounted for 46% of reduced rebates. Of the reduced rebates, 38% ($39 787 620) were attributable to 1 insulin product (Humalog; Eli Lilly); the additional rebate would have reduced the drug’s 2017 US revenues by 2.3% ($1717.8 million⁴). Ten of the top 25 drugs were manufactured by Bausch (formerly Valeant), with reduced rebates of $20 852 438 (20% of total reduced rebates). The reported revenues for 5 of these drugs⁵ indicated that additional rebates would have reduced revenues by 3.3%. Daraprim (Vyera Pharmaceuticals, formerly Turing), which increased prices by 5000% in 2015, had the second-largest amount ($5 763 297; 6% of total reduced rebates); additional rebates would have reduced Daraprim revenues by 8.7%.⁶

Table. Top 25 Brand-name Drugs by Reduced Rebate, Triggering the 2017 Medicaid Rebate Cap .

Brand Name	Generic Name	Reduced Rebate		Manufacturer
		US $	% of Total
Humalog 100 unit/mL	Insulin lispro	39 787 620	38	Eli Lilly
Daraprim 25 mg	Pyrimethamine	5 763 297	6	Vyera/Turing
Premarin Vaginal 0.625 mg/g	Estrogens, conjugated	5 703 122	6	Wyeth/Pfizer
Mephyton 5 mg	Phytonadione (Vit K1)	5 237 631	5	Bauscha
Syprine 250 mg	Trientine hydrochloride	4 867 126	5	Bauscha
Humulin R U-500 500 unit/mL	Insulin regular, Human	4 745 056	5	Eli Lilly
Cuprimine 250 mg	Penicillamine	3 449 201	3	Bauscha
Thiola 100 mg	Tiopronin	2 044 007	2	Bauscha
Renagel 800 mg	Sevelamer hydrochloride	1 619 155	2	Genzyme
Zovirax 5%	Acyclovir sodium	1 565 548	2	Bauscha
Duexis 800 mg-26.6 mg	Ibuprofen/famotidine	1 383 479	1	Horizon
Albenza 200 mg	Albendazole	1 236 365	1	Amedra
Humulin 70/30 100 unit/mL	Insulin Nph human isophane/insulin regular, human	1 101 176	1	Eli Lilly
Glumetza 1000 mg	Metformin hydrochloride	1 002 659	1	Bauscha
Rebif 44 μg/0.5 mL	Interferon beta-1a/albumin human	974 306	1	EMD Serono
Zonegran 100 mg	Zonisamide	924 751	1	Concordia
Felbatol 600 mg/5 mL	Felbamate	908 672	1	Meda/Mylan
Wellbutrin XL 300 mg	Bupropion hydrochloride	842 011	1	Bauscha
Felbatol 600 mg	Felbamate	738 465	1	Meda/Mylan
Rebif 22 μg/0.5 mL	Interferon beta-1a/albumin human	688 018	1	EMD Serono
Premarin 1.25 mg	Estrogens, conjugated	684 111	1	Wyeth/Pfizer
Glumetza 500 mg	Metformin hydrochloride	678 440	1	Bauscha
Novolin 100 unit/mL (70-30)	Insulin Nph human isophane/insulin regular, human	601 229	1	Novo Nordisk
Wellbutrin XL 150 mg	Bupropion hydrochloride	593 480	1	Bauscha
Demser 250 mg	Metyrosine	572 335	1	Bauscha
Duragesic 100 μg/h	Fentanyl	558 174	1	Janssen

^aOn July 17, 2018, Valeant changed its name to Bausch.

Discussion

Without the rebate cap, Medicaid expenditures would have been $103 million lower in 2017, representing 0.4% of 2016 net drug expenditures; drug-level revenues would be 2% to 9% lower for manufacturers that are subject to the highest additional rebates. Uncapped rebates would modestly decrease Medicaid expenditures, but manufacturer revenue reductions may be insufficient to discourage price increases. Increases are only deterred when the company would profit more from a smaller price increase that avoids the additional rebate rather than from a larger price increase that requires an additional rebate. Proposals to uncap rebates and add a rebate multiplier for large price increases may be a more effective deterrence.