Open debridement, irrigation with implant retention and antibiotic treatment (DAIR) is an accepted approach for early prosthetic joint infections (PJI). Our aim was to design a score to predict patients with a higher risk of failure.

From 1999 to 2014 early (<90 days) PJIs without signs of loosening of the prosthesis were treated with DAIR and were prospectively collected and retrospectively reviewed. The primary end-point was early failure defined as: 1) the need of an unscheduled surgery, 2) death-related infection within the first 60 days after debridement or 3) the need for suppressive antibiotic treatment. A score was built-up according to the logistic regression coefficients of variables available before debridement.

A total of 222 patients met the inclusion criteria. The most frequently isolated microorganisms were coagulase-negative staphylococci (95 cases, 42.8%) and Staphylococcus aureus (81 cases, 36.5%). Fifty-two (23.4%) cases failed. Independent predictors of failure were: chronic renal failure (OR:5.92, 95%CI:1.47–23.85), liver cirrhosis (OR:4.46, 95%CI:1.15–17.24), revision surgery (OR:4.34, 95%CI:1.34–14.04) or femoral neck fracture (OR:4.39, 95%CI:1.16–16.62) compared to primary arthroplasty, CRP >11.5 mg/dL (OR:12.308, 95%CI:4.56–33.19), cemented prosthesis (OR:8.71, 95%CI:1.95–38.97) and when all intraoperative cultures were positive (OR:6.30, 95%CI:1.84–21.53). Furthermore, CRP showed a direct relationship with the percentage of positive cultures (Linear equation, R2=0,046, P=0.002) and an inverse association with the time between the debridement and failure (Logarithmic equation, R2=0.179, P=0.003). A score for predicting the risk of failure was done using pre-operative factors (KLIC-score, figure 1) and it ranged between 0–9.5 points. Patients with a score ≤2, >2–3.5, 4–5, >5–6.5 and ≥7 had a failure rate of 4.5%, 19.4%, 55%, 71.4% and 100%, respectively.

The KLIC-score was highly predictive of early failure after debridement. In the future, it would be necessary to validate our score using cohorts from other institutions.

Early prosthetic joint infections (PJI) are managed with debridement, implant retention and antibiotics (DAIR). Our aim was to evaluate risk factors for failure after stopping antibiotic treatment.

From 1999 to 2013 early PJIs managed with DAIR were prospectively collected and retrospectively reviewed. The main variables potentially associated with outcome were gathered and the minimum follow-up was 2 years. Primary endpoint was implant removal or the need of reintroducing antibiotic treatment due to failure.

A total of 143 patients met the inclusion criteria. The failure rate after a median (IQR) duration of oral antibiotic treatment of 69 (45–95) days was 11.8%. In 92 cases PJI was due to gram-positive (GP) microorganisms, in 21 due to gram-negatives (GN) and 30 had a polymicrobial infection. In GP infections, combination of rifampin with linezolid, cotrimoxazole or clindamycin was associated with a higher failure rate (27.8%, P=0.026) in comparison to patients receiving a combination of rifampin with levofloxacin, ciprofloxacin or amoxicillin (8.3%) or monotherapy with linezolid or cotrimoxazole (0%) (Figure 1). Among patients with a GN infection, the use of fluoroquinolones was associated with a lower failure rate (7.1% vs 37.5%, P=0.044). Duration of antibiotic treatment was not associated with failure.

The only factor associated with failure was the oral antibiotic selection, but not the duration of treatment. Linezolid, cotrimoxazole and clindamycin but not levofloxacin serum concentrations are reduced by rifampin; a fact that could explain our findings. Further studies monitoring serum concentration could help to improve the efficacy of these antibiotics when combining with rifampin.