The Welfare Effects of Off-Label Prescriptions

Because of the high cost of securing a new indication for an existing drug from the Food and Drug Administration is very high, manufacturers often forgo modifications to their drugs’ labels, and off-label uses for approved drugs are common. However, little is known about which populations receive off-label prescriptions and whether the potential patient benefits from off-label prescriptions outweigh the costs. We combine nationally representative prescription data from the Medical Expenditure Panel Survey (MEPS) over the years 2007-2011 with data on Food and Drug Administration approved uses from the Physician’s Desk Reference (PDR). Our MEPS data contains information on what the drug was prescribed for as well as extensive demographic and insurance information. We match the MEPS data to PDR recorded indications to identify which prescriptions were for FDA-approved uses and which were off-label. We then model the likelihood that each drug was prescribed, accounting for clinical and socioeconomic characteristics of the patient, diagnoses, out-of-pocket costs, and whether the drug was on- or off-label using a Random Utility Model. Finally, the impact of off-label prescribing on patient welfare is assessed by simulating a counter-factual world where off-label prescribing is not allowed, and patients who have off-label drugs are re-assigned to the next-highest likelihood on-label option. The value in use (from a utility perspective) for patients under the actual world (with off-label prescribing) and the counter-factual world (no off-label prescribing) are then compared.