Article Text

Download PDFPDF

  1. L. LI1,2,
  2. N. Lu2,
  3. H. Xie2,3,
  4. J. Cibere1,2,
  5. J. Kopec1,2,
  6. J. Esdaile1,2,
  7. J. A. Aviña-Zubieta1,2
  1. 1The University of British Columbia, Vancouver, Canada
  2. 2Arthritis Research Canada, Richmond, Canada
  3. 3Simon Fraser University, Burnaby, Canada


Background: Both tramadol (narcotic-like drug) and nonsteroidal anti-inflammatory drugs (NSAIDs) are prescribed for pain relief among osteoarthritis (OA) patients. Evidence comparing risks of adverse events between tramadol and NSAIDs users is inconclusive.

Objectives: To examine the association of tramadol with all-cause mortality, cardiovascular disease (CVD), venous thromboembolism (VTE) and hip fractures (HFx) compared with NSAIDs and codeine in OA.

Methods: Design: Sequential propensity score-matched cohort study. Sample: All patients with OA who received medical care from 2005 to 2014 in the entire province of British Columbia, Canada. Tramadol cohort: Initial prescription of tramadol (n=56325). Four comparator cohorts: the initiation of one of the following: naproxen (n=13798), diclofenac (n=17675), cyclooxygenase-2 [Cox-2] inhibitor (n= 17039), or codeine (a weak opioid) (n=7813). Patients required to be prescribed neither tramadol nor its comparators during the year before the initial prescription date (i.e., index date). Outcomes: 1) all-cause mortality; first ever2) CVD, 3) VTE, 4) HFx within the 1st year after the initiation of tramadol or its comparators. Follow-up: from index date until the event occurred, disenrollment, or the end of a 1-year follow-up period. Statistical analysis: We created baseline covariates (demographics, comorbidities, medications and health resource utilization) from the year prior to the index date. Calendar years from 2005 to 2014 were divided into 10 blocks; propensity scores were calculated using logistic regression within each block. We used 1:1 greedy matching method. We estimated hazard ratios (HRs) using Cox proportional hazard models.

Results: After propensity score matching, 112650 patients with OA were included (mean age of 68 years, 62.8% were females). During the 1-year follow-up, 296 deaths (21.5/1000 person-years) occurred in the tramadol cohort and 246 (17.8/1000 person-years) in the naproxen cohort (Table 1). All-cause mortality was higher for tramadol compared with all NSAIDs cohorts, but not with the codeine cohort (Table 1, Figure 1). Tramadol initiators have also a higher risk of CVD and VTE compared with the diclofenac and Cox-2 inhibitor initiators with HRs ranging from 1.2 to 1.7. Furthermore, tramadol was also associated with a higher risk of HFx compared with all NSAIDs cohorts (HRs ranging from 1.4 to 1.5). No significant difference was found between tramadol and codeine (Table 1).

Table 1

Conclusion: OA patients initiating tramadol have an increased risk of mortality, CVD, VTE, and HFx within 1 year compared with NSAIDs, but no statistically significant difference in the risk was observed between tramadol and codeine.

Disclosure of Interests: None declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.