Comparing the Estimated Risk of Hip Fracture Among Subjects Exposed to Tramadol as Compared to Subjects Exposed to Codeine

Date of Approval: 
2020-09-01 00:00:00
Lay Summary: 
Hip fractures greatly impact an individual’s quality of life and carry a high risk of death within 1 year. Tramadol is a commonly used weak opioid for treatment of pain. A recent study by Wei et al. found that risk for hip fractures was higher for new users of tramadol than for new users of codeine. We had concerns in the way Wei et al. designed their study. We want to both duplicate their original work as well as run an improved analysis to see if we get similar results. We will execute these studies on CPRD (a similar database to what Wei et al. used) and additionally we will execute in three other US insurance claims databases to see if the results are also consistent across databases. The results from this study could help understand if people exposed to tramadol are at an increased risk of experiencing hip fractures when compared with individuals who take codeine.
Technical Summary: 
Hip fractures greatly impact an individual’s quality of life and carry a high risk of death within 1 year. Tramadol is a commonly used weak opioid for treatment of pain. A recent study by Wei et al. found that risk for hip fractures was higher for new users of tramadol than for new users of codeine. We had concerns in the way Wei et al. designed their study specifically. We plan to replicate the Wei et al. design as well as perform additional analysis that address the limitations – specifically creating tramadol and codeine cohorts that are more comparable. While the propensity score was not described in detail within Wei et al.’s work, we will implement large-scale propensity score fitting. Wei et al. also only used one data source, this study will use 4 databases total (one like what was used in the original work and 3 US claims datasets). Finally, for the comparisons being performed, negative controls will be implemented to check for residual confounding and adjust the p-value. We propose to do a study to assess hip fracture incidence among users of tramadol versus codeine that will reassess the relationship and address the Wei et al. study limitations.
Health Outcomes to be Measured: 
1 Outcome of Interest: Hip Fracture 101 Negative Control Outcomes (See Appendix A, some examples of the negative controls include epidermoid cyst, and neutropenia, tooth loss)
Application Number: 
20_173
Collaborators: 

Erica Voss - Chief Investigator - Johnson & Johnson ( JnJ - USA )
Erica Voss - Corresponding Applicant - Johnson & Johnson ( JnJ - USA )
Arun Singh - Collaborator - Janssen US
Daniel Fife - Collaborator - Johnson & Johnson ( JnJ - USA )
Gowtham Rao - Collaborator - Janssen US
Martijn Schuemie - Collaborator - Janssen US
Peter Rijnbeek - Collaborator - Erasmus University Medical Center ( EMC )
Rana Saberi - Collaborator - Janssen US