LAUSR

Abstract Objective This study sought to: (1) construct and validate a composite potential opioid misuse score; and (2) compare potential opioid misuse among individuals prescribed long-term therapy on tramadol, short-acting hydrocodone or short-acting oxycodone. Methods A retrospective cohort study was conducted using Arkansas All-Payer Claims Database (APCD; 2013–2018) linked to Arkansas Prescription Drug Monitoring Program (PDMP; 2014–2017) and state death certificate data (2013–2018). The study subjects were ambulatory, cancer-free adults with incident long-term therapy on tramadol, short-acting hydrocodone or short-acting oxycodone. The number of opioid prescribers/pharmacies, cash payment for opioid prescriptions, overlapping prescribers/pharmacies and a composite misuse score (derived from opioid prescribers/pharmacies and cash payment) were assessed in two 180 day windows as potential measures of misuse. The composite score was developed based on associations observed with opioid overdose and opioid-related injuries. Results A total of 17,816 (tramadol), 23,660 (hydrocodone) and 4799 (oxycodone) persons were included. The composite score had modest discrimination for overdose (c-index = 0.65). In the first 180 day period, the average composite misuse scores were 1.28 (tramadol), 1.93 (hydrocodone) and 2.18 (oxycodone). Compared to long-term hydrocodone, long-term tramadol had lower misuse (IRR [95% CI]: 0.75 [0.73–0.76]), and long-term oxycodone had higher misuse (1.09 [1.07–1.11]) in adjusted analyses. Qualitatively similar associations were observed for nearly all individual component measures of misuse. Conclusion A composite measure of potential opioid misuse had modest levels of discrimination in detecting overdose. In comparison to long-term hydrocodone therapy, long-term oxycodone had higher and tramadol had lower risk of potential opioid misuse.