ABSTRACT
PURPOSE: This study tests the hypothesis that there are no factors associated with removal or revision of a radial head prosthesis. A secondary analysis addressed the time to removal or revision. METHODS: We reviewed the database of two large hospitals from 2000 to 2014 and identified 278 patients that had radial head replacement after an acute fracture or fracture dislocation of the elbow: 19 had removal and 3 had revision of the radial head implant within the study period. Explanatory variables including demographics, the type of injury, prosthesis type, surgeon, medical centre, and associated injuries were evaluated. Survival analysis using Kaplan-Meier curves evaluated time to removal/revision. RESULTS: After adjustment for potential confounders using Cox regression multivariable analysis, hospital was the only factor independently associated with removal or revision (Hazard ratio=2.4, Confidence interval: 1.03-5.8, P value=0.043). The highest proportion of removal/revision was during the first year after implantation and decreased by half each year over the second to fourth years. The most common reason for removal of the prosthesis was to facilitate removal of heterotopic ossification (the majority with proximal radioulnar synostosis) rather than technical error or problems with the prostheses. CONCLUSION: These findings suggest that the decision to remove a radial head prosthesis may depend more on surgeon or hospital preferences than on objective problems with the prosthesis. Until clarified by additional study, removal of a prosthesis should not be considered an objective outcome in research. In addition, patients offered removal of a radial head prosthesis, might get the opinion of more than one surgeon at more than one hospital before deciding whether or not to proceed. LEVEL OF EVIDENCE: Level III Prognostic.
