Summary
Opioid harm can vary by opioid type. This observational study examined the effect of opioid type (oxycodone vs. tapentadol) on rates of persistent postoperative opioid use (‘persistence’). We linked hospital and community pharmacy data for surgical patients who were dispensed discharge opioids between 1 January 2016 and 30 September 2021. Patients were grouped by opioid experience (‘opioid‐naive’ having received no opioids in the 3 months before discharge) and formulation of discharge opioid (immediate release only or modified release ± immediate release). Mixed‐effects logistic regression models predicted persistence (continued use of any opioid at 90 days after discharge), controlling for key persistence risk factors. Of the 122,836 patients, 2.31% opioid‐naive and 27.24% opioid‐experienced patients met the criteria for persistence. For opioid‐naive patients receiving immediate release opioids, there was no significant effect of opioid type. Tapentadol modified release was associated with significantly lower odds of persistence compared with oxycodone modified release, OR (95%CI) 0.81 (0.69–0.94) for opioid‐naive patients and 0.81 (0.71–0.93) for opioid‐experienced patients. Among patients who underwent orthopaedic surgery (n = 19,832), regardless of opioid experience or opioid formulation, the odds of persistence were significantly lower for those who received tapentadol compared with oxycodone. This was one of the largest and most extensive studies of persistent postoperative opioid use, and the first that specifically examined persistence with tapentadol. There appeared to be lower odds of persistence for tapentadol compared with oxycodone among key subgroups, including patients prescribed modified release opioids and those undergoing orthopaedic surgery.