• Journal Article

Using stated preference modeling to forecast the effect of medication attributes on prescriptions of alcoholism medications

Citation

Swait, J., & Mark, T. (2003). Using stated preference modeling to forecast the effect of medication attributes on prescriptions of alcoholism medications. Value in Health, 6(4), 474-482. DOI: 10.1046/j.1524-4733.2003.64247.x

Abstract

Objective: The objective of this study was to forecast physicians' preferred rate of prescriptions of alcoholism medications given different medications attributes (i.e., efficacy, side effects, compliance, price, mode of administration, method of action).

Methods: Stated preference modeling was used. Data came from a survey of 1388 physicians specializing in addiction medication (65% response rate). Physicians were given four hypothetical scenarios, each in which they were asked to choose between prescribing one of two hypothetical alcoholism medications with given attributes or prescribing no medication.

Results: Prescribing decisions were elastic with respect to the efficacy of the medication (1.35 and 1.65 using two efficacy measures). A 10% increase in the percentage of patients who would remain abstinent on the medication would lead to a 13.5% increase in the percentage of patients prescribed the medication. Prescribing decisions were inelastic with respect of nonserious side effects (-0.24), compliance (0.80), and price (-0.25). The market share of alcoholism medications with extremely favorable characteristics (i.e., 80% abstinence rate, a 95% no heavy drinking rate, a 10% side effect rate, a 80% compliance rate, and a price of $0.25) was predicted to be 53%, and 47% of the population would not be prescribed a medication to prevent alcoholism.

Conclusions: The market share of new medications to treat alcoholism among addiction specialists could surpass the low usage rates of existing medications if those medications have better attributes. However, prescription levels may not reach that expected for treatment of other diseases.