Outcomes Associated with Treatment of Chronic Pain with Tapentadol Compared with Morphine and Oxycodone: A UK Primary Care Observational Study

INTRODUCTION: This study compared adverse outcomes and resource use for patients with a diagnosis of pain treated with tapentadol prolonged-release (PR) versus those treated with morphine controlled-release (CR) and oxycodone CR. METHODS: Data were sourced from the Clinical Practice Research Datalink (CPRD), a database derived from UK primary care. Patients prescribed tapentadol PR between May 2011 and December 2016 were selected and matched to two groups of controls treated with either morphine CR or oxycodone CR on gender, age, pain duration, pain site, pain aetiology, Charlson index and prior analgesia. Times to first adverse event (constipation or nausea/vomiting) were compared within a Cox proportional hazards model. Rates of primary care contacts, accident and emergency contacts and, for a subset of patients linked to Hospital Episode Statistics (HES), inpatient admissions and outpatient contacts were compared using incidence rate ratios (IRRs) derived from Poisson regression. RESULTS: A total of 1907 patients prescribed tapentadol PR were identified and 1791 (93.9%) had a pain diagnosis. Of these 1246 (65.3%) were matched to morphine controls and 829 (43.4%) to oxycodone controls. Compared to controls, gastrointestinal adverse events with tapentadol PR treatment were reduced; aHR = 0.532 (0.402–0.703; p < 0.001) versus morphine CR and 0.517 (0.363–0.735; p < 0.001) versus oxycodone CR. Compared with morphine CR, primary care contacts [IRR = 0.831 (0.802–0.861)], accident and emergency attendance [0.739 (0.572–0.951)], outpatient contacts [0.917 (0.851–0.989)] and inpatients contacts [0.789 (0.664–0.938)] were reduced. For oxycodone, the respective figures were 0.735 (0.703–0.768), 0.971 (0.699–1.352), 0.877 (0.799–0.962) and 0.748 (0.601–0.932). CONCLUSION: Tapentadol PR was associated with significantly fewer adverse gastrointestinal events than morphine CR and oxycodone CR in patients with a diagnosis of pain. There was also significantly reduced primary and secondary care resource use. As with all observational studies, potential bias due to residual confounding and confounding by indication should be considered. FUNDING: Grünenthal Ltd. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s12325-019-00932-7) contains supplementary material, which is available to authorized users.