Does Cost-Sharing Affect Response to Signals? The Effect of Bad News on Utilization of Prescription Drugs with Cost-Sharing

Background: Rising health care costs have increasingly led to pharmaceutical benefit plans with greater consumer cost-sharing. Evaluations of such plans suggest that increased cost sharing has had the intended impact of reducing consumer spending on prescription drugs. However, it is not clear how cost-sharing will affect consumer response to new information on drugs. The pharmaceutical market is constantly flooded with new information. New drug related information—positive or negative—often starts emerging after a drug's approval. On the negative side it could be release of research results pointing to toxicity or other problems, leading in some cases, to Food and Drug Administration (FDA) drug risk communication such as a black box warning. The impact of such information is complicated due to the presence of different levels of cost-sharing. This paper examines the joint effect of new information and consumer cost-sharing on the consumption of prescription drugs.

Methods: In this paper I use MarketScan database of privately insured individuals in the United States. The database has information on 3.5 million individuals covered by employer sponsored health insurance plans. It captures person-specific clinical utilization, expenditures, and enrollment across inpatient, outpatient, prescription drug. For purposes of this paper, I focus on individuals over the age of 18 with pharmaceutical claims information. I analyze four different files belonging to the database for the period 2007-2010. These are 1) The Outpatient Pharmaceutical Claims Table, which contains the insurance drug claims for all individuals who purchased prescription drugs, 2) Enrollment Detail Table, which linked individuals to their health plan enrollment history. 3) Inpatient Admissions Table, and the 4) Outpatient Services Table containing information on patients’ medical conditions and certain demographic characteristics. To measure new information I focus on FDA's drug risk communications, that are categorized into three levels depending on risk severity. Using difference-in-difference analysis I identify the impact of new information and cost sharing on pharmaceutical drug consumption.

Discussion: Given different levels of cost sharing, we should expect the consumer response to any negative information regarding a particular drug to be slower at lower levels of cost-sharing. However, as a competing hypothesis one would expect that the response to a severe drug risk communication to be swift and not responsive to cost sharing. The evidence on the impact of publicity and direct-to-consumer advertisement on prescription drug demand has found that not all forms of publicity are equal. This paper extends this line of research by examining how different levels of cost-sharing and different types of drug risk communication interact to influence consumer response.