Tuesday, November 11, 2014

New Walgreens Data Verifies That 340B Reduces Retail Generic Dispensing Rates

November’s Health Affairs includes The 340B Discount Program: Outpatient Prescription Dispensing Patterns Through Contract Pharmacies In 2012. (Sorry, the full article is available only to subscribers.)

The paper inadvertently provides compelling evidence that the 340B Drug Pricing Program’s crazy incentives are lowering generic dispensing rates at retail pharmacies. The two lead authors work at Walgreens, which perhaps explains why they try (unsuccessfully) to spin the results away from this unfortunate conclusion.

With few exceptions, third-party payers and PBMs have been minimally engaged in the public policy discussion about how the 340B program should be modernized. Walgreens’ unexpected disclosures about cost-shifting to managed care suggest that they should start paying attention. Read on and see if you agree.


The chart below shows the generic dispensing rate (GDR) for the top five therapeutic categories with 340B prescriptions. In all cases, the GDR is statistically significantly lower for 340B prescriptions when compared with all prescriptions (p < 0.0001 for all comparisons).

[Click to Enlarge]

For covered entities, the greatest 340B rebates come from brand-name drugs. The Walgreens data support the idea that 340B entities encourage prescribing brand-name drugs when therapeutically equivalent generic alternatives exist. In other words, the 340B program reduces generic dispensing rate at contract pharmacies.

Here’s why this matters:
  • Influence over Pharmacies—By aligning with 340B entities, contract pharmacies are paid by them—rather than earning traditional dispensing spreads and fees. Third-party payers may not realize it, but the 340B program is hijacking their pharmacy networks, undermining their benefit design strategies, and generating extra profits for hospitals by raising payers’ costs.
  • Cost-Shifting to Managed Care—The majority of 340B prescriptions are paid by Medicare, Medicaid, and commercial third-party payers. For example, Senator Charles Grassley’s inquiries have revealed that only 1 in 20 patients served by Duke University Health System’s 340B pharmacy was uninsured. The remaining 95 percent’s prescription costs were paid by Medicare, Medicaid, and private insurance. See Hospitals' Extraordinary 340B Pharmacy Profits from Insured Patients
To "follow the 340B dollar" and track see exactly how the money flows, see Exhibit 97 (page 163) in the 2013–14 Economic Report on Retail, Mail, and Specialty Pharmacies.

The article’s authors offer no clear explanation for the disparity in generic dispensing rates shown above. Here’s the best they can muster:
“[T]he specific medications dispensed in a therapeutic class may differ between 340B dispensing and overall dispensing because of variations in comorbidity and underlying severity of conditions between the two patient populations.”
While this sentiment may explain utilization differences between therapeutic classes, it’s hardly compelling for explaining differences within a single therapeutic class.


I’m slightly skeptical about the authors’ claim that 340B prescriptions accounted for less than 0.5% of all Walgreens prescriptions.

The paper states:
  • "The sample included approximately 500 million dispensed prescriptions for patients of all ages and with all types of insurance coverage in the fifty states, the District of Columbia, and Puerto Rico."
  • "340B prescriptions amounted to less than 0.5 percent of the approximately 500 million prescriptions dispensed by Walgreens in 2012."
However, Walgreens SEC filings show that the company filled many more than 500 million prescriptions.
  • Walgreens' FY2012 10-K (for the 12 months ending 8/31/12) states: "Overall, Walgreens filled approximately 664 million prescriptions in 2012...Adjusted to 30-day equivalents, prescriptions filled were 784 million in fiscal 2012..."
  • Walgreen's FY2013 10-K (for the 12 months ending 8/31/13) states: "Overall, Walgreens filled approximately 683 million prescriptions in fiscal 2013...Adjusted to 30-day equivalents, prescriptions filled were 821 million in fiscal 2013..."
Walgreens confirmed to me that the analysis was based on actual—not equivalent—prescriptions. So, about 170 million prescriptions were eliminated from the sample. The study’s authors told me:
"The study was based on a statistically valid sample of approximately 500 million prescriptions. The study criteria were equally applied to 340B and non-340B prescriptions."
I’m still curious to know more about which prescriptions were excluded from the analysis and how these exclusions changed the results. Since 2012, the number of contract pharmacies has grown by 60%, implying that even the questionable 0.5% figure would be higher.


In 2013, Senator Grassley asked Walgreens for data about the drugstore chain’s profits from the 340B drug discount program. See Senator Grassley Grills Walgreens About Its 340B Profits. If the senator has received a response, it has yet to be made public.

Senator Grassley is likely to head the Senate Judiciary Committee in the new Republican-led Congress. I expect this Health Affairs article to feature prominently in his next request for details.


  1. The reason is pretty simple. As one who is smarter than me stated; "follow the money". Economics and profit will always drive the engine. It would be nice if a review would provide insights from an Econ 101 point of view.

  2. Adam, good information, but not a huge surprise. With the 340B discounts, it's very possible that some brand name drugs have a lower net cost to the contract pharmacy and 340B covered entity than their generic equivalents. And, it's also possible that the dispensing fee to the pharmacy on a 340B prescription (as paid by the covered entity) is higher than it would be than if the prescription were filled under a patient's prescription benefit (if applicable). Thanks for another (as always) insightful article.

  3. In today's article you,say the majority of 340B Rx's are paid, among others, by Medicaid. However my understanding is that, under most circumstances, Mediciad Rx's are excluded from the 340B program, as has been my experience for the past 10+ years.

  4. According to the three North Carolina hospitals' disclosures to Senator Grassley, Medicaid paid for 7% to 18% of 340B prescriptions. See the table in Hospitals' Extraordinary 340B Pharmacy Profits from Insured Patients, which comes from page 3 of Senator Grassley's letter to HRSA.

  5. Walgreens typically exlcudes FFS Medicaid prescriptions from their 340B agreements so I would anticipate very few 340B Medicaid prescriptions in this assessment.
    Only a subset of prescriptions that are 340B eligible are pulled into the program by Walgreens (many pharmacies just pull in brand items). Inexpensive generics with little to no savings created by 340B pricing are not "converted" to 340B in many contract pharmacy agreements. It's not surprising that the "converted" prescriptions are more heavily weighted as brand / specialty.
    The 340B program incentivizes contract pharmacy relationships with entities that have a large specialty population (oral oncology, HIV, HCV) as the 340B discounts are often well above the ceiling 23% for these drugs. The savings are significant. It's not at all surprising that a focus on HIV and other specialty populations leads to the brand:generic disparities. For example, the antiviral category for 340B will be heavily weighted towards HIV (expensive brands) while the general population will see a big dose of generic acyclovir and valacyclovir thrown in.

    I'm not aware of any hospital of clinic that asks providers to select brand agents over generic agents to increase 340B discounts. This would be unethical, would likely be resisted heaviliy by physicians, and would eventually come to light and cause significant embarassment for the hospital or clinic leadership.

    Nothing much to see here folks. I have no idea why Walgreens published this and I have no idea what significance it has in the market.