Comparative Risks of Fractures and Mortality in Patients with Osteoarthritis initiating Tramadol versus Other Opioids

Background: Tramadol is frequently prescribed to treat pain in patients with osteoarthritis (OA). Little is known, however, about its comparative safety in this high-risk population.

Objectives: To evaluate the risks of fracture and all-cause mortality in OA patients who initiated tramadol vs other prescription opioids.

Methods: Using the Optum de-identified Clinformatics® Data Mart Database (2005-2021), we identified individuals aged > 50 years who initiated tramadol, codeine, oxycodone, or hydrocodone following at least 365 days of enrollment in the database. Patients were required to have an OA diagnosis recorded during the year before opioid initiation. Propensity-score matching weights were used to balance demographics, comorbidities, and healthcare utilization during baseline. Outcomes included fractures, and all-cause mortality. We conducted both as-treated analysis, with patients censored on outcome occurrence, death, disenrollment, hip or knee replacement, index opioid discontinuation, initiation of another opioid, or end of study period, and a 60-day intention to treat (ITT) analysis, with patients followed for 60 days regardless treatment changes. Cox proportional hazard models estimated hazard ratios (HR) and 95% confidence intervals (CI), with tramadol as reference.

Results: We identified 1,059,127 eligible individuals with OA who initiated one of the study opioids. Weighted population included 83,038 tramadol, 83,405 codeine, 82,638 hydrocode, and 83,350 oxycodone individuals. As-treated analysis of fracture yielded HRs of 0.93 (95% CI 0.81 – 1.06) for codeine; 1.09 (95% CI 1.00 – 1.18) for hydrocodone; and 1.20 (95% CI 1.07 – 1.33) for oxycodone. The 60-day ITT analysis yielded HRs of 0.90 (95% CI 0.82 – 0.99) for codeine; 0.98 (95% CI 0.98 – 1.04) for hydrocodone; and 1.04 (95% CI 0.96 – 1.13) for oxycodone. The as-treated analysis of mortality yielded HRs of 1.09 (95% CI 0.85 – 1.38) for codeine, 1.26 (95% CI 1.11 – 1.44) for hydrocodone, and 2.59 (95% CI 2.19 – 3.05) for oxycodone. The 60-day mortality HRs in ITT analyses were 0.75 (95% CI 0.67 – 0.85) for codeine, 0.99 (95% CI 0.92 – 1.07) for hydrocodone, and 1.63 (95% CI 1.49 – 1.78) for oxycodone.

Conclusions: In this study, we observed that OA patients who initiated tramadol had lower rates of fracture and mortality than patients initiating oxycodone or hydrocodone in as-treated analysis; though no difference between patients initiating hydrocodone and tramadol was observed in ITT analysis. Patients who initiated tramadol had higher rates of fracture and mortality than patients on codeine in ITT analysis, but not in as-treated analysis.