A Prescription for Manipulation? Impact of the 340B Drug Discount Program on Hospitals

The 340B drug discount program substantially reduces the acquisition cost of outpatient drugs for hospitals that provide a sufficient amount of inpatient care to Medicaid and low-income Medicare patients. The program provides strong incentives for hospitals to alter patient mix in order to gain entry to the program. The goals of our analysis are two-fold. First, we test for manipulation and estimate changes in patient and service line mix in anticipation of gaining eligibility to the program. Second, we estimate the causal impact for a sub-set of hospitals that did not manipulate entry into the program. To test for manipulation we use Medicare hospital cost report data from 1987-2016 and McCrary Density Tests to estimate the extent of manipulation of the program eligibility criterion around the eligibility threshold. We find strong evidence of manipulation for hospitals after 2003. Event study analyses reveal that hospitals that began participating after 2003 increased provision of labor and delivery services and decreased provision of psychiatric care up to 4 periods before becoming eligible for 340B. To estimate the impact of the 340B program for a sample of non-manipulating hospitals, we combine a difference-in-differences approach with inverse probability of treatment re-weighting to estimate the causal effect of the program on hospitals that began participating in 2003 or earlier. We find that 340B decreases outpatient drug costs, and increases outpatient revenue and outpatient drug charges, with little effect on inpatient drug charges. The 340B program did not increase hospital profitability, perhaps because participant hospitals were more likely to provide unreimbursed care as well as offer several low-profit services lines after participating in the. Our results suggest 340B participants comprise a diverse group of hospitals and that recently proposed cuts to the program may have varying effects.