Adopting Average Acquisition Cost – Are Drugs Now Cheaper for Medicaid?

This paper addresses how drug pricing transparency initiatives affect public spending on pharmaceuticals by looking at the effects of adopting so-called Actual (average) Acquisition Cost (AAC) reimbursement rules by state Medicaid agencies.

Medicaid pharmaceutical reimbursement is seeing one of its most dramatic institutional changes in a generation. Average Wholesale Price (AWP) and Wholesale Acquisition Cost (WAC) long served as the primary benchmarks that Medicaid agencies use to guide their payments to pharmacies and other providers for prescription drugs. However, detractors have long argued that manufacturers control the AWP and WAC, and can exploit a lack of transparency about the “actual” prices of their products to drive excessive payments for their products. Starting with Alabama in 2011, several states adopted alternative, ostensibly more transparent, bases for pharmaceutical reimbursement based on surveys to derive estimates of pharmacy AACs. This system of reimbursement has now been enshrined in the 2016 Covered Outpatient Drug final rule (2/1/2016) requiring all states to establish aggregate upper limits on drug spending based on AAC plus dispensing fees.

As a matter of theory these institutional changes alone should not have a static effect on overall pharmaceutical spending because pharmacy participation in state Medicaid networks is voluntary. Pharmacies that participate in Medicaid opt into expected flows of patients, prescriptions, and Medicaid payments with associated costs and benefits. State-level Medicaid pharmacy payments therefore should reflect an equilibrium tradeoff between the generousness of the Medicaid contract and the extent of pharmacy participation. Changes in contract mechanics alone (such as the drug payment basis) are unlikely to affect the overall economics required to ensure adequate pharmacy participation. A material reduction in overall Medicaid payments to pharmacies would likely result in a drop in pharmacy participation over time. If, on the other hand, rents were being generated non-transparent drug prices, then lowering overall reimbursement and reverting to normal economic returns should not affect overall participation.

This study presents an empirical examination of the effect of state-level adoption of AAC reimbursement on overall Medicaid drug expenditures. We rely on the Medicaid State Drug Utilization Dataset (SDUD) for prices and quantity weights over time. Using both interrupted time series models for fixed-weight product baskets, and difference-in-differences regressions, we find that per-prescription spending did not generally decrease in states that adopted AAC reimbursement. Moreover, we find that spending on prescriptions for lower priced tiers of drugs (i.e., generics) generally increased in states that adopted AAC, while spending decreased for higher-priced (branded) drugs. Much of this shift appears due to the increased dispensing fees adopted as part of the change to AAC-based reimbursement. We also explore the dynamic impacts of these compositional changes in reimbursement.