BACKGROUND:
The specific risk factors for surgical site infection (SSI) in orthopaedic oncology patients undergoing endoprosthetic reconstruction have not previously been evaluated in a large prospective cohort. In the current study, we aimed to define patient- and procedure-specific risk factors for SSI in patients who underwent surgical excision and endoprosthetic reconstruction for lower-extremity bone or soft-tissue tumors using the prospectively collected data of the Prophylactic Antibiotic Regimens in Tumor Surgery (PARITY) trial.

METHODS:
PARITY was a multicenter, blinded, randomized controlled trial with a parallel 2-arm design that aimed to determine the effect of a long duration (5 days) versus short duration (24 hours) of postoperative prophylactic antibiotics on the rate of SSI in patients undergoing surgical excision and endoprosthetic reconstruction of the femur or tibia. In this secondary analysis of the PARITY data, a multivariate Cox proportional hazards regression model was constructed to explore predictors of SSI within 1 year postoperatively.

RESULTS:
A total of 96 (15.9%) of the 604 patients experienced an SSI. Of the 23 variables analyzed in the univariate analysis, 4 variables achieved significance: preoperative diagnosis, operative time, volume of muscle excised, and hospital length of stay (LOS). However, only hospital LOS was found to be independently predictive of SSI in the multivariate regression analysis (hazard ratio per day = 1.03; 95% confidence interval = 1.01 to 1.05; p < 0.001). An omnibus test of model coefficients demonstrated that the model showed significant improvement over the null model (χ2 = 78.04; p < 0.001). No multicollinearity was found.

CONCLUSIONS:
This secondary analysis of the PARITY study data found that the only independent risk factor for SSI on multivariate analysis was hospital LOS. It may therefore be reasonable for clinicians to consider streamlined discharge plans for orthopaedic oncology patients to potentially reduce the risk of SSI.

LEVEL OF EVIDENCE:
Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.