Abstract
Aim
Treatment of complicated wound healing after total joint arthroplasty is controversial. What exactly constitutes prolonged wound drainage is matter of debate and recommendations to manage it vary considerably. Nonoperative measures are often recommended. If drainage persists, surgery may be indicated. To further intricate decision-making, differentiating superficial from deep surgical site infection is also controversial and inherently complex. Specific cutoffs for synovial fluid leukocyte count and blood C-reactive protein (CRP) in the acute stage have been suggested as a way to superficial infection requiring superficial wound washout from deep infection requiring a formal debridement, antibiotics and implant retention (DAIR) procedure. The goal of this study is to analyze clinical and laboratory findings of an institutional protocol of “aggressively” proceeding with formal DAIR in all patients with complicated wound healing
Method
Our indications for DAIR in suspected acute postoperative periprosthetic joint infection (PJI) are: 1)prolonged wound drainage and CRP upward trend after day-3; 2)persistent wound drainage by day-10 regardless of CRP; 3)wound healing disturbance (e.g. “superficial” infection, “superficial” skin necrosis) anytime in early postoperative weeks. We retrospectively evaluated patients undergoing DAIR in the first 60 postoperative days between 2014–2018. Patients without multiple deep tissue cultures obtained intraoperative were excluded. Deep infection was defined by at least two positive deep tissue cultures or one positive deep culture and positive leukocyte count (>10,000 cells/mL or >90% PMN).
Results
A total of 44 DAIR procedures were included. Deep infection was confirmed in 79.5%(35/44) of cases. Mean CRP in infected cases was 93mg/L with 63%(19/30) of them below the 100 mg/L threshold. Unfortunately, only a small proportion of cases (10/44) had synovial fluid leukocyte counts available. Mean leukocyte count was 15,558 cells/mL and mean proportion of PMN was 65.3%. Of these ten, six confirmed deep infections were below the proposed >10,000 cells/mL or >90% PMN cutoff.
Conclusions
Early diagnosis of acute postoperative PJI is often hampered by its very subtle presentation. This study confirms that more often than not, deep infection is present when facing complicated wound healing after total joint arthroplasty, supporting our institutional “aggressive” protocol. We have been surprised by the number of confirmed acute PJI with low blood CRP levels and low synovial leukocyte counts. We hypothesize that the proposed acute PJI specific thresholds may lead to misinterpret a significant proportion of cases as superficial infections thus compromising timely intervention. The findings of this study lack confirmation in larger cohorts
