Failure analysis of infection persistence after septic revision surgery: a checklist algorithm for risk factors in knee and hip arthroplasty

Purpose Several studies describe risk factors for primary periprosthetic joint infection (PJI) and general treatment outcome factors like microbe spectrum or patient-specific risk factors. However, these general and patient dependent findings cannot solely explain all cases of infection persistence after a prior septic revision. This study analyzes possible specific and patient independent reasons for failure after revisions for PJI in knee and hip arthroplasty. Methods In a prospective analysis all patients were included that were treated: (1) at our department, (2) with a two-stage exchange, (3) between 2013 and 2017, (4) due to an infection persistence after a previous revision for PJI. Possible reasons for infection persistence were identified using a checklist algorithm, based on international guidelines. Results 70 patients with infection persistence could be included (44 knee joints, 26 hip joints). The average age was 71 years, the CCI (Charlson Comorbidity Index) 2.8 and the ASA (American Society of Anesthesiologists) score 2.7. In 85% at least one possible reason for patient independent infection persistence could be identified analyzing the previous infection therapy: (1) 50% inadequate therapy concept (n = 35), (2) 33% inadequate surgical debridement (n = 23), (3) 30% inadequate antimicrobial therapy (n = 21), (4) 13% missed external bacterial primary focus (n = 9). After the individual failure analysis, all 70 patients were treated with a two-stage exchange in our department and in 94.9% infection freedom could be achieved (34.3 +/- 10.9 months follow-up). Conclusions In the majority of failed revisions with subsequent infection persistence at least one possible patient independent failure cause could be identified. The entire previous therapy should be critically reviewed following failing revisions to optimize the outcome of septic revisions. By using a checklist algorithm, high rates of infection freedom were achieved.