The 340B Program Hits a Crisis Point
As the Centers for Medicare & Medicaid Services (CMS) implement nearly $1.6 billion in cuts to the 340B Drug Pricing Program, a new study looks at the consequences of the program, and questions whether it has had its intended effect of helping safety-net hospitals serve poor and vulnerable populations.
In the New England Journal of Medicine, Penn LDI PhD alumna Sunita Desai, now an Assistant Professor at NYU, and Harvard’s Michael McWilliams conclude that the 340B program has been associated with increased hospital-physician consolidation in hematology-oncology, and outpatient hospital administration of certain drugs in hematology-oncology and ophthalmology, without clear evidence of expanded care for low-income patients.
The 340B Drug Pricing Program was created in 1992 to increase outpatient access to prescription drugs and necessary services for underserved populations by stretching providers’ federal resources. It was created as a response to the shortcomings of the Medicaid Drug Rebate Program of 1990. This program mandated Medicaid discounts for drugs, and had the unintended effect of pharmaceutical manufacturers discontinuing charitable discounts to safety net providers to ensure that average prices remained high.
340B requires pharmaceutical manufacturers to allow eligible providers – disproportionate share hospitals (DSHs) that serve low-income patient populations – to purchase outpatient drugs at a steep discount. MEDPAC estimated that the average discount was 22.5% of the average sales price (ASP). The eligibility criteria for 340B has been a point of contention: general acute hospitals must have at least an 11.75% DSH percentage, which is largely based on the proportion of inpatient (and not outpatient) services delivered to low-income populations. The program has also grown significantly – from 591 participating hospitals in 2005 to 1,673 (nearly a third of all hospitals in the U.S.) in 2011.
It is important to note that 340B hospitals get the drug discount for their entire patient populations, and they are still reimbursed for the full cost of the drugs (until January 1, 2018 as discussed below), which makes drug administration more profitable for them. One of the critiques of 340B is that hospitals are then incentivized to administer more drugs, especially to their privately insured population, from whom they get the greatest ‘profit’. The findings of the Desai and McWilliams study seem to give this argument some clout.
It is not surprising that 340B providers have balked at CMS’s cuts to the program - instead of reimbursing hospitals at the ASP plus 6%, Medicare will now reimburse at ASP minus 22.5%, though reimbursement for other services were increased as an offset. For more than half of hospitals, reductions in Medicare Part B drug revenue are estimated to be less than 5%, though about 6% of providers could see cuts exceeding 10%.
The resources generated from 340B discounts are meant to help the hospital better serve its low-income population, but the program has no direct requirement that it do so. The language of the law suggests that Congress meant to indirectly provide health care organizations with resources to foster cross-subsidization of unprofitable care. Hospitals claim that they use their savings from the program to expand services to vulnerable populations, which the current cuts would undermine. 340B hospitals provide about 60% of all uncompensated care in the U.S., and there is some evidence that they do serve more low-income patients than non-340B hospitals, though this could be attributable to factors unrelated to 340B participation (e.g., differences in local patient populations).
The Desai and McWilliams NEJM study contributes to the available evidence by looking at associations between financial gains for 340B hospitals and expanded care for low-income patients. The authors were also interested in the effect of 340B on hospital-physician consolidation, and on the outpatient administration of drugs in hospital-owned facilities, which would speak to misaligned incentives of the program. For their analysis, they looked at non-profit general acute care hospitals with 50 or more beds, and with a DSH percentage within 10 percentage points of the 11.75% eligibility threshold.
Their findings suggest that hospitals eligible for 340B have responded to program incentives by increasing the outpatient provision of parenteral drugs, and in the case of hematology-oncology, by acquiring physician practices. They describe changes in the payer mix as well:
Our findings also suggest the program prompted eligible hospitals to treat more Medicare patients who are more likely to have private supplemental insurance to cover the 20% of Part B drug costs that is not covered by Medicare. The finding that patients served by eligible hospitals were less likely to have Medicaid, which reimburses hospitals less generously than other forms of supplemental insurance, is consistent with the financial incentives of hospitals and with evidence that 340B-participating hospitals have increasingly affiliated with hematology-oncology practices serving affluent communities.
The authors found no evidence of hospitals using the additional resources to invest in safety-net providers or provide more inpatient care to low-income populations. There are a number of important policy implications of these findings. One highlighted by the authors is that "policies intended to improve or expand care for medically underserved populations may be ineffective if they rely on indirect mechanisms with weak incentives, such as the cross-subsidization that the 340B program intends for hospitals to implement.”
The Trump Administration’s cuts to 340B have pitted the pharmaceutical industry against 340B hospitals. Not surprisingly the Desai and McWilliams study raised the ire of the American Hospital Association (AHA). Austin Frakt has a ‘point and counterpoint’ post about this on The Incidental Economist, which points to the AHA’s methodological critiques as well as Desai and McWilliams’ response. They reject the suggestion that the study should be dismissed due to some of its limitations, saying that it “presents strong evidence of hospital responses to Program incentives that are inconsistent with Program goals, evidence that is sufficient to motivate Program reform and more generally inform the design of policies intended to enhance care for low-income populations.” Whether this evidence does motivate reform of 340B remains to be seen.