S1E1: What is Causing Rising Drug Launch Prices? (Prof. David Ridley)
In today’s episode, we discuss drug launch prices and how they are impacted by the way that Medicare reimburses providers for them.
My guest is Professor David Ridley, faculty director for Health Sector Management at Duke University’s Fuqua School of Business. He was the lead author of a paper that proposed the priority review voucher program, which became law in 2007 and created a market of over one billion dollars for drug development for neglected diseases.
The paper we discuss can be found here.
The YouTube mentioned in the episode that summarizes Ridley’s paper is below:
Professor Ridley’s website about the Priority Review Voucher Program is here
In-depth show notes
00:36 – 03:15. Introductions to the case study of Aduhelm (aducanumab) and our guest Dr. David Ridley.
03:15 – 09:05. Background on how healthcare providers are reimbursed.
- The Medicare Modernization Act of 2003 (AKA the Medicare Prescription Drug, Improvement, and Modernization Act) changed the way Medicare Part B reimburses providers from AWP to ASP.
- Wholesale Acquisition Cost: price that a wholesaler pays a manufacturer for a drug
- Average Wholesale Price (AWP): an elevated sticker price that was often higher than actual market prices or the Wholesale Acquisition Cost, therefore given the nickname “Ain’t What’s Paid”; how Medicare Part B reimbursed providers prior to 2005
- Average Sales Price (ASP): reimbursement based on past prices; how Medicare Part B reimburses providers now, post-Medicare Modernization Act
- Medicare Part B: covers medical costs in the outpatient setting, including clinic visits, home health, emergency department services, or ambulatory surgeries; medications covered under Medicare Part B are usually those patients would not administer to themselves and would receive at a doctor’s office or other outpatient setting (e.g. chemotherapy, intravenous immunoglobulins, injectable drugs that meet certain criteria)
09:06 – 17:20. Dr. Ridley’s paper used AnalySource data from 1999-2010 and the difference-in-differences and interrupted time series analyses to answer the question: Does reimbursement based on past prices cause launch prices to change?
- AnalySource: database containing decades of many types of drug prices (including WAC, ASP, drug launch prices) across many doses and manufacturers; does not include drug quantities
- Difference in differences: a common, “quasi-experimental” research design in economics that emulates the randomized control trial
- Interrupted time series: another analysis technique that looks at trends over time; compares trends before and after an intervention begins
17:20 – 30:19. We discuss some challenges, unexpected results, and other methodological considerations along the way.
- Hypotheses in Dr. Ridley’s paper that were not supported by the data
- Challenges of conducting health economics research
- Randomized controlled trials vs quasi-experimental methods
- Economics/social science literature vs medical literature
- Importance of replicating results to show internal validity
- Placebo tests to look for the absence of an effect when an intervention has not taken place yet
30:20 – 44:26. The analyses supported their hypothesis that reimbursement based on the past price of the drug caused launch prices to rise dramatically. This negative consequence of reimbursing based on past prices, they propose an alternative model of reimbursement.
- Reimbursement could be made based on the cost-effectiveness of the drug, i.e. cost per Quality-Adjusted Life Year (QALY) added, similar to the UK’s drug-pricing model
- Reimbursement based on cost-effectiveness could still lead to unintended incentives (e.g. lower drug spending could lead to lower investment in R&D)
- Strategies for de-risking drug development for pharmaceutical companies
- Importance of placing protections (i.e. placing high value, potentially overpaying) for drugs for rare diseases, which have the potential to be undervalued in a reimbursement model based on cost-effectiveness but serve as a form of insurance
- Full versus fair insurance parameters
- Returning to the example of Aduhelm; what Aduhelm pricing would look like in Dr. Ridley’s cost-effectiveness model and how that may impact research incentives
44:27 – 56:19. Finally, we discuss Dr. Ridley’s 2006 proposal for a Priority Review Voucher (PRV) to incentivize the development of tropical diseases and expedite the development of drugs for other diseases
- How the PRV program works and became law in 2007
- Changes that were made between Dr. Ridley’s original proposal and the final FDA policy
- PRV expansion to include rare pediatric conditions in 2012
- Effects of the PRV program since it became law
- How to get academic work out into the real world
References
- https://www.cms.gov/newsroom/press-releases/cms-finalizes-medicare-coverage-policy-monoclonal-antibodies-directed-against-amyloid-treatment
- https://www.nytimes.com/2022/04/07/health/aduhelm-medicare-alzheimers.html
- https://www.cms.gov/medicare-coverage-database/view/ncacal-decision-memo.aspx?proposed=N&ncaid=305
- https://www.healthaffairs.org/do/10.1377/forefront.20210622.349716/full/
- https://www.hhs.gov/answers/medicare-and-medicaid/what-is-medicare-part-c/index.html
- https://faculty.fuqua.duke.edu/~dbr1/research/Medicare-drug-prices.pdf
- https://www.medicare.gov/sites/default/files/2018-09/11435-Are-You-an-Inpatient-or-Outpatient.pdf
- https://www.congress.gov/bill/108th-congress/house-bill/1
- https://dimewiki.worldbank.org/Difference-in-Differences
- https://diff.healthpolicydatascience.org/#introduction
- https://www.unicef-irc.org/KM/IE/img/downloads/Quasi-Experimental_Design_and_Methods_ENG.pdf
- https://scholar.princeton.edu/sites/default/files/deaton/files/deaton_randomization_revisited_v2_2019_01.pdf
- https://youtu.be/UB1A62u9fBE
- https://www.povertyactionlab.org/resource/introduction-randomized-evaluations
- https://www.medpac.gov/improving-medicares-payment-for-part-b-drugs-requiring-pharmaceutical-manufacturer-reporting-of-sales-price-data/#:~:text=Medicare%20pays%20providers%20106%20percent,actually%20pays%20for%20the%20product.
