Periprosthetic infection (PPI) is one of the most devastating complications of total knee arthroplasty (TKA). It is widely accepted that resection arthroplasty supplemented with intravenous antibiotics and delayed exchange arthroplasty is the treatment modality of choice for infected TKA. However, the outcome after reimplantation has varied and unpredictable results have been reported. This study evaluates the outcome of this treatment strategy in a single high volume specialised center. Furthermore, our study aims to identify the factors that lead to failure of this treatment.

A thorough review of our joint registry database revealed that 80 patients with an infected TKA underwent resection arthroplasty at our institution during 2000–2005. Sixty-five patients underwent two-stage exchange arthroplasty while the remaining 15 failed to have the second stage reimplantation due to ill health or underwent arthrodesis or amputation. The latter 15 were excluded from the analysis. All patients were followed up prospectively for at least two years. Detailed data including demographics, comorbidities, surgical history, and medication intake was collected. Intraoperative data, organism profile, and complications were also documented. Failure was defined as patient requiring additional surgical procedure for control of infection or loosening.

Two-stage exchange arthroplasty successfully eradicated infection in 45 patients (31%) without need for further treatment. Twelve patients (18%) had recurrent infection that necessitated another resection arthroplasty. Eleven (17%) patients required irrigation and debridement for postoperative purulent drainage which successfully treated infection in 5 cases (46%). The remaining 6 patients failed and required resection arthroplasty. Three additional patients had early loosening of components and required revision arthroplasty. The exact cause of loosening in these patients could not be determined, and despite lack of isolation of organisms infection was suspected. Our analysis identified that irrigation and debridement prior to resection arthroplasty are major risk factors for failure.

Current strategies to treat periprosthetic infection remain imperfect. Two-stage exchange arthroplasty with all its inherent problems and inconveniences imparted a modest success in treatment of PPI at our high volume specialised center. The rise in the number of resistant and virulent organisms, increase in the number of patients with severe medical comorbidities who develop infection may account for the decline in the success of two-stage resection arthroplasty. Novel strategies for treatment of PPI are desperately needed.