The Price Elasticity of Specialty Drug Use: Evidence from Cancer Patients in Medicare Part D

Monday, June 13, 2016: 10:55 AM
G60 (Huntsman Hall)

Author(s): Jeah Kyoungrae Jung; Roger Feldman; Marshall McBean

Discussant: Erin A. Taylor

We estimated how sensitive patients are to specialty drug benefits in cancer drug use among elderly Medicare Part D enrollees. Medicare Part D is delivered through private plans with different benefit schemes. We exploited variation in drug benefit generosity across Part D plans and estimated the price responsiveness in specialty cancer drug use. We utilized Part D plan formulary data, which became available recently in a format linkable to patient-level data, to create new measures of specialty drug cost-sharing during the initial coverage phase (i.e., prior to the gap). We computed a plan-level expected out-of-pocket (OOP) spending index as a weighted average of the initial copayment for all specialty cancer drugs, where the weight is the share of each drug’s fills of the total specialty cancer drug fills in the entire sample. This approach improves on prior research that used actual drug prices, which in fact reflect patients’ choices influenced by drug benefits and thus distort comparisons of benefit generosity across plans. We corrected potential endogeneity associated with drug benefits due to voluntary Part D plan choice by instrumenting OOP indexes.

The study population was a random sample of Medicare Fee-For-Service beneficiaries with cancer between 2010 and 2012. We used 2-stage residual inclusion (2SRI) estimation to analyze the specialty drug use, and used a 2-state Least Squares (2SLS) method to analyze annual total specialty cancer drug spending among specialty drug users. Our analysis suggests that Medicare beneficiaries with cancer are responsive to specialty drug prices: the probability of using a specialty cancer drug during the year decreases by 0.04 percentage points for a $100 increase in the plan’s initial monthly copayment. This corresponds to about a 3 percent increase in the use of a specialty cancer drug. The estimated price elasticity at the mean copayment was -0.37. However, we find that among those who use a specialty drug(s), total specialty cancer drug spending is not responsive to cost-sharing. We found similar results from several additional analyses including analysis of four common cancer types (lung, breast, prostate, and colon), and analysis of the top Part D specialty drug (Gleevec) among leukemia patients. These findings indicate that even patients who suffer from a complex and debilitating condition such as cancer consider price as an important factor in seeking treatment. While this suggests that cost-sharing could be an effective tool to manage specialty drug utilization, high cost-sharing could be a significant barrier to access to needed drugs for patients who would receive substantial benefits from those treatments. It would be important to identify circumstances or patients’ clinical states where specialty cancer drugs can be most valuable and to ensure access to treatments for such patients.