Abstract
Aim
Periprosthetic joint infection (PJI) is a feared complication of total joint arthroplasty of hip (THA) or knee (TKA). Debridement, antibiotic treatment, and implant retention (DAIR) is an effective treatment of early PJI. In the Netherlands, cefazolin resistance in early PJI after primary arthroplasty is low. Little is known about causative micro-organisms and resistance patterns in PJI after revision arthroplasty. No recommendations for empirical treatment are described in the current guidelines. The aim of this study is to describe the characteristics of PJI after revision arthroplasty and to evaluate whether the used empirical treatment regimens are adequate, based on microbiology data.
Method
In this retrospective study we included patients with early PJI after aseptic revision of THA or TKA, treated with DAIR between 2012 and 2020. Success rate was defined as implant retention and no persistent or recurrent infection during one year follow-up.
Results
We identified 96 patients with PJI. PJI was most frequently caused by Staphylococcus spp. (n=73), Gram-negative bacilli (n=31) or Enterococcus spp. (n=13). Polymicrobial infection was diagnosed in 38 PJIs. Mismatches were present in 72 (75%) of the PJIs (95% CI: 0.66–0.84). Table 1 shows the number of mismatches per empirical treatment regimen. Figure 1 shows the responsible micro-organisms for the mismatches. Success rate of PJI treatment was significant reduced for patients with mismatching compared to matching empirical therapy: 62% vs. 95% respectively (OR: 0.09, 95% CI: 0.01–0.68, p=0.004). If vancomycin would have been the empirical treatment, mismatches would have been reduced to 31 (32%) (95% CI: 0.23–0.42). With vancomycin-ciprofloxacin combination therapy the mismatches would have been reduced to 1% (95% CI: −0.01–0.03).
Conclusions
There is a high number of mismatches in empirical treatment in early PJI after revision arthroplasty, which have significant influence on the outcome. Based on our data cefazolin should not be recommended as empirical treatment for this specific group. Our data shows that review of local data is necessary to improve treatment strategies, that eventually might improve outcome. Besides changing Gram-positive coverage, a prospective study is needed to assess the benefits of broader spectrum empiric antimicrobial treatment taken into account toxicity and other side effects such as antimicrobial resistance.
For any tables or figures, please contact the authors directly.
