Decomposition of Expenditures for Insulins Across the Distribution Chain—Insights Into Rising Prices?

There is little debate that increasing drug costs are a substantial concern for patients and payers. Per capita spending on prescription drugs in the US is the highest in the world.¹ In a 2019 poll of US adults, 79% of respondents believed that costs of prescription drugs are unreasonable, 24% indicated difficulty affording their medications, and nearly 1 in 5 indicated that they did not fill a prescription in the past year because of the cost.²

What is debatable is why drug costs are increasing in the US. Most explanations for increasing drug costs start with drug companies, and indeed they are an easy target. Manufacturers in the US are allowed to set their own price for new products, which are then protected from competition through patent exclusivity.¹ Manufacturers have been able to extend those patents, in many cases for years beyond the original patent expiration.¹ During that exclusivity period, prices continue to increase, with a 2020 report³ documenting that list prices of branded drugs increased at a yearly average of 9.1% from 2007 to 2018. Thus, at face value, manufacturers would seem to be largely responsible for rising drug costs, by setting high prices for new drugs entering the market, limiting competition through patent extensions, and raising prices on existing brand-name products.

However, increasing list prices of drugs are partially mediated by ever-increasing discounts. Thus, drug manufacturers point to the net price of drugs (the cost of drugs after all discounts have been applied) rather than the list price as representing a more accurate figure of drug costs. Although the lack of transparency regarding discount data does not allow for direct analysis of the association of discounts with net prices of drugs, previous analyses indicate that discounts on branded products accounted for 60% the increase in drug costs from 2007 to 2018.³ One type of discount that has received foremost attention because of its magnitude is the pharmacy benefit manager (PBM) rebate. The PBM rebates are discounts negotiated between insurers and manufacturers through PBMs. These rebates are retrospectively paid based on volume of purchases and linked to sales at the patient level.¹ Because patient out-of-pocket expenses are based on prerebate drug prices, patients do not directly benefit from these discounts. Indeed, increasing rebates (tied to increased list prices) result in higher patient copays, potentially increasing profits of PBMs.⁴ This potential misalignment of incentives has led some to target PBMs for their role in increasing drug costs. In response, PBMs argue that their negotiation power results in deep discounts of popular drugs through rebates, effectively lowering net costs by passing along most of the rebates to health plans, which allows them to control premiums.

There are other agents involved in the drug distribution chain and payment system in the US⁴ which could also bear in the responsibility for drug price inflation. Pharmacies (including specialty pharmacies) and drug wholesalers/distributors also participate in the drug distribution chain and receive payment for the provision of drugs and additional services to patients. On the demand side of the drug distribution and reimbursement chain, there are payers/insurers and patients.⁴ These entities are responsible for paying for drugs and thus do not receive a net flow of money from purchasing drugs (flow of money is negative after discounts, because they pay for drugs). Given this situation, patients and payers/insurers desire the lowest drug spending and have no incentive to increase drug costs.

Recent research^3,5 has presented analyses to determine the spread between list and net prices, but to our knowledge, there are no published reports which have documented the net flow of money across the entire drug distribution system. The American Diabetes Association formed a Working Group to gain greater understanding of drug pricing in general and insulin pricing in particular.⁶ They attempted to follow money across the insulin supply chain, but after substantial research and stakeholder discussions, they were unable to determine the money flow across the system and the profits earned by each intermediary.⁶ Along with many others, they point to the lack of transparency and complexity of the entire supply chain as the reason that making reliable estimates of the distribution of costs for each entity is not possible.

More insight into the relative role of each entity in the drug supply chain in the overall cost of drugs would be immensely important in understanding factors associated with increasing drug prices and informing health policy. Within this context, in this issue of JAMA Health Forum Van Nuys and colleagues⁷ present an analysis of changes from 2014 to 2018 in the flow of money across the insulin distribution system. They focus on insulin for several reasons: insulins are lifesaving, and increasing list prices of these products have been the focus of intense media, academic, and governmental criticism.⁷ High insulin list prices and high patient out-of-pocket costs have resulted in harm to patients,⁷ with some patients rationing doses and risking adverse clinical outcomes. In addition, their analysis using insulin was facilitated by several state drug price transparency initiatives that provided information on spread payments and other fees paid to PBMs.

Van Nuys and colleagues⁷ report that list prices, as expected, increased (by 40.1%), but net prices decreased (by 30.8%). These estimates are consistent with prior findings (37% decrease in net prices in 2014-2018).³ A particularly notable result is the substantial increase in the net flow of money to pharmacies, accounting for more than 20% of overall spend in 2018. This estimate could be partially explained by the inclusion of 340-B discounts under the net flow of money to pharmacies using the authors’ methods. We are unaware of any reports documenting such a substantial increase in compensation to pharmacies to dispense insulins. Although it may be counterintuitive that health plans (insurers) are included in net flow expenditures because they along with patients are the ultimate payers, presumably this figure represents the proportion of the rebate money that is returned to insurers.

The methods that Van Nuys et al⁷ used to calculate their figures are complex. They used the SSR Health Rx Brand Pricing Data to obtain list and net prices. However, the authors also rely on Drug Channels Institute data for a number of inputs in the analysis, and the lack of independent validation of this source is an important limitation. For other costs, they used data from various sources including commercial claims data sets for weighting of use of insulin products, Medicare Part D claims for patient out-of-pocket expenses, Medicaid data for spread payments, a Nevada legislative report for retention of rebates for essential diabetes medications (not insulin directly), and Securities and Exchange Commission filings to determine wholesale margins for all drugs (not insulin individually). We agree with the authors in their acknowledgment that the use of different patient populations and of non–insulin-specific data are a substantial limitation in their calculations. We also concur with them that the combination of data sets representing different populations is often necessary in drug pricing research because of the lack of transparency in the money flow in the drug reimbursement system. Policy makers desire sound estimates of drug prices and money flows to base decisions; sound estimates will only be possible if state or federal authorities require greater transparency in the transactions along the drug supply chain.

Despite these limitations, the authors are to be congratulated for attempting such a daunting task. We hope that other investigators will follow their lead in attempting to piece together more information about the flow of money through the drug distribution system, and that others will seek to confirm their findings.

Published: November 5, 2021. doi:10.1001/jamahealthforum.2021.2258

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Good CB et al. JAMA Health Forum.

Corresponding Author: Chester B. Good, MD, MPH, Insurance Services Division, UPMC Health Plan, 600 Grant St, Pittsburgh, PA 15219 (goodcb2@upmc.edu).

Conflict of Interest Disclosures: Dr Good reported receiving funding for the Center for Value-Based Pharmacy Initiatives, which he directs from Evernorth Financial outside the submitted work; and being an employee of UPMC Health Plan, Insurance Services Division. Dr Hernandez reported receiving personal fees from Pfizer and personal fees from Bristol-Myers Squibb outside the submitted work.