ABSTRACT
OBJECTIVES: This study compared the risks of opioid shopping behavior and opioid abuse between tapentadol immediate release and oxycodone immediate release and, to validate the definition of shopping, examined the association between opioid shopping and opioid abuse further. MATERIALS AND METHODS: This retrospective cohort study using linked dispensing and diagnosis databases followed opioid-naive patients for development of shopping behavior and/or opioid abuse during 1 year after initial exposure to tapentadol or oxycodone. Shopping was defined by having overlapping opioid prescriptions from >1 prescriber filled at ≥3 pharmacies; abuse by having International Classification of Diseases, 9th revision diagnoses reflecting opioid abuse, addiction, or dependence. To determine their association, we cross-tabulated shopping and opioid abuse and calculated odds ratios. Risks of developing each outcome were estimated using logistic regression. RESULTS: Among 277,401 participants initiating opioid use with tapentadol (39,524) or oxycodone (237,877), 0.6% developed shopping behavior, 0.75% developed abuse. Higher proportions of patients in the oxycodone group developed shopping behavior and abuse than in the tapentadol group (shopping: adjusted odds ratio [95% confidence interval], 0.45 [0.36-0.55]; abuse: 0.44 [0.37-0.54]). Shopping behavior and abuse were associated; of those with shopping behavior, 6.5% had abuse. Age (18 to 64 y), sex (male), prior benzodiazepine use, paying cash, and history (mood disorders, abuse of nonopioid medications, and back pain) were risk factors for developing either outcome. DISCUSSION: Shopping behavior and abuse measure complementary, but associated, constructs, which further validates the current definition of shopping. The risk of developing either is lower among patients who initiate opioid use with tapentadol than those who initiate opioid use with oxycodone.