Quantifying the Value of Personalized Medicines: Evidence from COX-2 Inhibitors

Abstract We develop a conceptual framework for estimating the value of personalized medicines. We show that personalizing medicines generates value from two sources. The first is a market-expansion effect by persons who initiate treatment due to reduced pre-treatment uncertainty about the effectiveness or side effects of treatment. The second is a market-contraction effect due to discontinuation of treatment by persons unresponsive to treatment. We apply the conceptual framework to evaluate the value of a predictive test to assess whether patients are at elevated risk for cardiac complications from COX-2 inhibitors. We find that this predictive test would yield an overall value to patients of about $16 billion per year or $1284 per likely patient.