Response to Letter to the Editor on “Rivaroxaban Use for Thromboprophylaxis Is Associated With Early Periprosthetic Joint Infection”

Abstract

To the Editor: We would like to thank the authors for their input regarding our article entitled “Rivaroxaban Use for Thrombosis Prophylaxis Is Associated With Early Periprosthetic Joint Infection” [1]. As the authors noted, in the RECORD randomized controlled trials, no statistical differences in safety were reported [2, 3, 4, 5]. However, the adverse events that were well described in the methodology were primarily limited to bleeding events. There was no description of postoperative wound infection, how this was defined, or for what time point this was measured. In the absence of these details, we assumed that these data were reported for the treatment period, which was 14 days postoperative. This is a diversion from the methodology used in our study, in which patients were followed postoperatively for 30 days. In summary, although we agree with the authors that these trials did not show any statistical differences in terms of infection, we feel that this may have been a cursory look at the wrong time point for this outcome. In response to the comment regarding potential confounders contributing to our results (eg, steroid use, comorbidities, history of smoking, and so forth), we were not able to statistically adjust for these factors owing to the low incidence of early deep surgical site infections in our cohort as we noted in our limitations. However, we did perform univariate comparisons of the 2 treatment groups, and no statistically significant differences were observed between the 2 groups. We have discontinued the use of rivaroxaban at our institution, as it is our opinion that its benefits do not outweigh its risks.