Post-operative prophylactic antibiotics in aseptic revision hip and knee arthroplasty: a propensity score matching analysis

Abstract

AbstractThe use of extended antibiotic (EA) prophylaxis (> 24 h) remains controversial in aseptic revision arthroplasty. We sought to determine whether EA prophylaxis reduces the risk of periprosthetic joint infection (PJI) in aseptic revision hip and knee arthroplasty. A total of 2800 patients undergoing aseptic revision hip and knee arthroplasty at five institutional databases from 2008 to 2017 were evaluated. One to two nearest-neighbor propensity score matching analysis was conducted between patients who did and did not receive extended antibiotic prophylaxis. The matching elements included age, sex, body mass index, Charlson comorbidity index, hospital distribution, year of surgery, joint (hip or knee), surgical time, CRP, preoperative hemoglobin, albumin, and length of stay. The primary outcome was the development of PJI, which was assessed at 30 days, 90 days, and 1 year following revision and analyzed separately. A total of 2467 (88%) patients received EA prophylaxis, and 333 (12%) patients received standard antibiotic (SA) prophylaxis (≤ 24 h). In the propensity-matched analysis, there was no difference between patients who received EA prophylaxis and those who did not in terms of 30-day PJI (0.3% vs. 0.3%, p = 1.00), 90-day PJI (1.7% vs. 2.1%, p = 0.62) and 1- year PJI (3.8% vs. 6.0%, p = 0.109). For revision hip, the incidence of PJI was 0.2% vs 0% at 30 days (p = 0.482), 1.6% vs 1.4% at 90 days (p = 0.837), and 3.4% vs 5.1% at 1 year (p = 0.305) in the EA and SA group. For revision knee, the incidence of PJI was 0.4% vs 0.9% at 30 days (p = 0.63), 1.8% vs 3.4% at 90 days (p = 0.331), and 4.4% vs 7.8% at 1 year (p = 0.203) in the EA and SA group. A post hoc power analysis revealed an adequate sample size with a beta value of 83%. In addition, the risks of Clostridium difficile and resistant organism infection were not increased. This multi-institutional study demonstrated no difference in the rate of PJIs between patients who received extended antibiotic prophylaxis and those who did not in aseptic revision arthroplasty. The risk of C. difficile and resistant organism infection was not increased with prolonged antibiotic use.