Factors Associated with Choice and Success of One- Versus Two-Stage Revision Arthroplasty for Infected Hip and Knee Prostheses

Abstract

Background

Prosthetic joint infections (PJI) of hip and knee arthroplasties are becoming increasingly common with an aging population and an increasing demand for these procedures. Despite attempts at standardization, medical and surgical approaches vary widely among practitioners.

Questions/Purposes

We first sought to determine if there were specific factors associated with choice of one- versus two-stage revision. Then we investigated whether the type of revision approach influenced clinical treatment success. Finally, among two-stage revisions, we assessed if an antibiotic holiday prior to the second procedure affected clinical treatment success.

Methods

We retrospectively reviewed patients who had revision surgery for infection of a hip or knee arthroplasty between January 1, 2000, and December 31, 2013, at the Sunnybrook Health Sciences Centre and the Holland Orthopedic and Arthritic Centre. PJI cases were identified using the Ontario Joint Replacement Registry. Infection was defined by gross intraoperative evidence of infection, positive intraoperative culture(s), and/or sinus tract prior to operation. The primary outcome was treatment failure at 1 year after revision surgery based on requirement for further surgery, ongoing infection, and/or continuous suppressive antibiotics.

Results

Of 110 eligible patients identified, 35 patients had a one-stage and 75 patients had a two-stage revision. Choice of a one-stage approach was most influenced by particular surgeon preference and was more likely for hip revision versus knee revision (OR 3.39 (95%CI 1.85–6.23). There was no statistical difference in clinical treatment success rate between one-stage (33/35; 94.2% success) and two-stage revision (63/75; 84%; p = 0.13). Enterococcus spp. (21 versus 3%; p = 0.027) and Peptostreptococcus spp. (14 versus 1%; p = 0.042) were more common among clinical treatment failures than successes, with a trend towards the same for Staphylococcus aureus (29 versus 9%; p = 0.06). Additionally, treatment success was not influenced by whether the patient had an antibiotic holiday with a two-stage revision.

Conclusion

Our findings confirm the uncertainty of surgical strategy for treatment of PJI in hip and knee arthroplasty. Superiority of one- versus two-stage revision and the value of antibiotic-free periods prior to definitive revision remain unclear. Large prospective studies or randomized controlled trials are needed to inform best practice for treatment of these complex clinical problems.