Development and validation of an Opioid Attractiveness Scale: a novel measure of the attractiveness of opioid products to potential abusers

BACKGROUND The growing trends in opioid abuse, assessment of the abuse liability of prescription opioid products, and growing efforts by the pharmaceutical industry to develop 'abuse-resistant' formulations highlight a need to understand the features that make one product more 'attractive' than another to potential abusers. We developed a scale to measure the 'attractiveness' of prescription opioids to potential abusers, and used the scale to measure the relative attractiveness of 14 opioid analgesic products. METHODS First, the concept of attractiveness was empirically defined with a group of prescription opioid abusers and experts in opioid abuse using a process called Concept Mapping. Abuse liability consisted of two components: factors intrinsic to the drug formulation (e.g., speed of onset, duration) and factors extrinsic to drug formulation (e.g., availability, availability of alternatives, cost). A 17-item Opioid Attractiveness Scale (OAS) was constructed, focusing on factors intrinsic to the drug product. RESULTS A total of 144 individuals participated in tests of validity and reliability. Internal consistency was excellent (Cronbach's alpha = 0.85-0.94). Drug rankings based on OAS scores achieved good inter-rater agreement (Kendall's W 0.37, p < 0.001). Agreement on drug OAS scores between the developmental sample and a confirmation sample was good (IntraClass Correlations [ICC] of 0.65-0.69). Global ratings of overall attractiveness of the 14 selected opioid products by substance abuse counselors corresponded with the rankings based on OAS ratings of the abuser group. Finally, substance abuse counselors completed the OAS, yielding a high level of correspondence with ratings by the abuser group (ICC = 0.83, p = 0.002). The OAS differentiated attractiveness among 14 selected pharmaceutical opioid products. OxyContin, Dilaudid, and Percocet were ranked highest (most attractive); Talwin NX and Duragesic were ranked lowest (least attractive). CONCLUSION An initial examination of the psychometric properties of the OAS suggests that it is a valid and reliable scale. The OAS may be useful in providing important guidance on product features that are attractive to potential abusers.