College of Psychology: Faculty Articles

Title

Measuring Attractiveness for Abuse of Prescription Opioids

Document Type

Article

Publication Date

1-2010

Keywords

Opioid Analgesics, 80 and over, Pharmaceutical Chemistry, Statistical Data Interpretation, Demography, Drug Costs, Health Surveys, Internet, Psychological Models, Opioid-Related Disorders, Pain, Prescription Drugs

Publication Title

Pain Medicine

ISSN

1526-2375

Volume

11

Issue/No.

1

Abstract

Objective: Prescription opioids are the second most misused/abused drug in the United States behind only marijuana. Recreational prescription opioid users appear to prefer some products over others; however, the extent to which attributes of any particular formulation account for such preferences has yet to be determined. The Opioid Attractiveness Technology Scaling was developed to identify the particular features of a prescription opioid that are relevant to its attractiveness for recreational use, and to use these features to model attractiveness for recreational use of particular prescription opioid formulations.

Design: Four hundred and ninety-one self-reported recreational prescription opioid users identified 43 product features as being relevant to determining whether a product is “attractive” or “unattractive” for recreational use. Average ratings were used to determine appropriate weights to be applied to the features. A factor analysis yielded 10, highly differentiated factors. Five hundred and sixty-four prescription opioid abusers were then asked to rate the extent to which the 43 features identified in Study 1 were relevant to specific prescription opioid products they had used.

Results: Respondents provided an overall preference rating of these products and a model was created. A random intercept model yielded a significant pseudo R2 of 0.14 (chi-square = 310.02, degrees of freedom [df] = 10, P < 0.001). The model fit least well, albeit significantly, for abusers who preferred to swallow the drug (pseudo R2 = 0.06; chi-square = 55.52, df = 10, P < 0.001) and best for those who preferred to inject (pseudo R2 = 0.37; chi-square = 199.34, df = 10, P < 0.001).

Conclusions: The relevance of the model is discussed along with possible modifications that might allow prediction of “attractiveness” of “abuse deterrent” formulations that have not yet been marketed.

DOI

10.1111/j.1526-4637.2009.00736.x