key: cord-0018222-fns0ogpz
authors: Yeung, Kai; Dusetzina, Stacie B.; Basu, Anirban
title: Association of Branded Prescription Drug Rebate Size and Patient Out-of-Pocket Costs in a Nationally Representative Sample, 2007-2018
date: 2021-06-14
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2021.13393
sha: 2ac17864247fa2b0bc35d840bfccacb2452e6ea6
doc_id: 18222
cord_uid: fns0ogpz

IMPORTANCE: Over the past decade, branded prescription drug manufacturers have substantially increased list prices while offering larger rebate payments to health care insurers. Whereas larger rebates can partially offset increases in list prices for insurers, patient out-of-pocket costs may be directly associated with list prices for individuals without insurance and indirectly associated with list prices for individuals with insurance through deductibles or coinsurance. OBJECTIVE: To investigate the association between rebates and patient out-of-pocket costs and whether this association differs by coverage type (ie, Medicare, commercial, or uninsured) and before and after 2014. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study was conducted using data from the Medical Expenditure Panel Survey (MEPS) combined with pricing data for single-source branded drugs from SSR Health from 2007 through 2018. The study was conducted among a nationally representative sample of the noninstitutionalized civilian US population. Included individuals were respondents to MEPS with at least 1 prescription for a single-source branded drug who were covered by Medicare or commercial insurance or were uninsured during an entire year. Data analyses were conducted from August 2019 through March 2021. EXPOSURES: Estimated rebate size. MAIN OUTCOMES AND MEASURES: Out-of-pocket costs per prescription were calculated, adjusting for year and drug. RESULTS: Among 38 131 individuals with at least 1 prescription, the mean age was 54 years (95% CI, 54 to 55 years), with 22 044 women (57.8%) and 29 086 White individuals (76.3%). The sample included 444 unique drugs with a survey-weighted total of 4.7 billion prescriptions. Estimated mean (SE) rebates increased from $34 ($1) per prescription in 2007 to $374 ($9) per prescription in 2018. The rebate sizes were associated with statistically significant mean out-of-pocket increases per branded prescription of $4 (95% CI, $4 to $4) from 2007 to 2013 and $11 (95% CI, $10 to $12) from 2014 to 2018. From 2014 to 2018, rebate sizes were associated with statistically significant mean increases in out-of-pocket costs per prescription of $13 (95% CI, $12 to $13) for individuals with Medicare, $6 (95% CI, $6 to $7) for individuals with commercial insurance, and $39 (95% CI, $34 to $44) for individuals without insurance. After adjusting for list prices, there was no association between rebates and out-of-pocket costs, with a change from 2014 to 2018 of −$0.01 (95% CI, −$0.04 to $0.02). CONCLUSIONS AND RELEVANCE: These findings suggest that drug manufacturers may have provided larger rebates to insurers primarily by increasing list prices and that individuals without insurance had greater cost increases. The results emphasize the need for policy solutions that decouple list prices and out-of-pocket costs.

To generate our overall estimates, we used a linear regression model according to the following specification:

[ | • ]) = 0 + 1 + 2 _ + 3 + 4 + 5 + 6 * _ + 7 * + 8 _ * + 9 * _ * +

Where:

Y icdt = Out-of-pocket costs in 2018 US dollars for prescription fill "i" for drug "d" in year "t"

Rebate idt = Estimated rebate size in 2018 US dollars for prescription fill "i" for drug "d" in year "t" To generate our estimates by coverage type, we used the same model specification, except that we excluded the "cov_type c " term and repeated the regressions in each of the 3 subpopulations (Medicare, Commercial, and Uninsured). 

38 to 47.57) Prolensa 14.08 (-10.92 to 39.08) Prolia 64

98 (-30.96 to 62.92) Revatio

Commercial*Rebate = Commercial times Rebate) CI, confidence interval eReferences

Changes in list prices, net prices, and discounts for branded drugs in the US

 Table 2 0.01 (-0.04 to 0.05) 0.15 (0.00 to 0.31) a Presented as elasticity estimates: percent change in out-of-pocket costs associated with 1% change in estimated rebates b Probability of incurring out-of-pocket costs: i.e., extensive margin c Increase in out-of-pocket costs, given non-zero out-of-pocket costs: i.e., intensive margin CI, confidence interval