Non-Cefazolin Antibiotic Prophylaxis is Associated with Higher Rates of Elbow Periprosthetic Joint Infection

Introduction Periprosthetic joint infection (PJI) is a common source of failure following elbow arthroplasty. Perioperative prophylactic antibiotics are considered standard of care. However, there are no data regarding the comparative efficacy of various antibiotics in the prevention of PJI for elbow arthroplasty. Previous studies in shoulder, hip, and knee arthroplasty have demonstrated higher rates of PJI with administration of non-cefazolin antibiotics. The elbow has higher rates of PJI than other joints. Therefore, this study evaluated whether perioperative antibiotic choice affects rates of PJI in elbow arthroplasty. Materials & Methods A single institution prospectively collected Total Joint Registry database was queried to identify patients who underwent primary elbow arthroplasty between 2003 and 2021. Elbows with known infection prior to arthroplasty (25) and procedures with incomplete perioperative antibiotic data (7) were excluded, for a final sample size of 603 total elbow arthroplasties and 19 distal humerus hemiarthroplasties. Cefazolin was administered in 561 elbows (90%) and non-cefazolin antibiotics including vancomycin (32 elbows, 5%), clindamycin (27 elbows, 4%) and piperacillin/tazobactam (2 elbows, 0.3%) were administered in the remaining 61 elbows (10%). Univariate and multivariate analyses were conducted to determine the association between the antibiotic administered and the development of PJI. Infection-free survivorship was estimated using the Kaplan-Meier (KM) method. Results Deep infection occurred in 47 elbows (7.5%) and 16 elbows (2.5%) were diagnosed with superficial infections. Univariate analysis demonstrated that patients receiving non-cefazolin alternatives were at significantly higher risk for any infection (Hazard Ratio (HR) 2.6, 95% confidence interval [CI] 1.4-5.0]; p < 0.01) and deep infection (HR 2.7 [95% CI 1.3 – 5.5]; p < 0.01) compared with cefazolin administration. Multivariable analysis, controlling for several independent predictors of PJI (tobacco use, male sex, surgical indication other than osteoarthritis, and American Society of Anesthesiologists score), showed that non-cefazolin administration had a higher risk for any infection (HR 2.8 [CI 1.4 – 5.3]; p < 0.01) and deep infection (HR 2.9 [95% CI 1.3 – 6.3]; p < 0.01). Survivorship free of infection was significantly higher at all time points for the cefazolin cohort (Figure 1). Discussion In primary elbow arthroplasty, cefazolin administration was associated with significantly lower rates of PJI compared to non-cefazolin antibiotics, even in patients with a greater number of prior surgeries which is known to increase the risk of PJI. For patients with penicillin or cephalosporin allergies, preoperative allergy testing or a cefazolin test dose should be considered prior to administering non-cefazolin alternatives.