RESUMO
BACKGROUND: We hypothesized, that periprosthetic joint infection (PJI) accounts for the major proportion of first (primary) and repeated (secondary) Total Knee Arthroplasty revisions at our university referral arthroplasty center. METHODS: One thousand one hundred forty-three revisions, performed between 2008 and 2016 were grouped into primary (55%) and secondary (45%) revisions. The rate of revision indications was calculated and indications were categorized by time after index operation. The odds ratios of the indications for primary versus secondary revision were calculated. RESULTS: In the primary revision group PJI accounted for 22.3%, instability for 20.0%, aseptic loosening for 14.9% and retropatellar arthrosis for 14.2%. PJI (25.6%) was the most common indication up to 1 year after implantation, retropatellar arthrosis (26.8%) 1-3 years and aseptic loosening (25.6%) more than 3 years after implantation. In the secondary revision group PJI accounted for 39.7%, aseptic loosening for 16.2% and instability for 13.2%. PJI was the most common indication at any time of revision with 43.8% up to one, 35.4% 1-3 years and 39.4% more the 3 years after index operation. The odds ratios in repeated revision were 2.32 times higher (p = 0.000) for PJI. For instability and retropatellar arthrosis the odds ratios were 0.60 times (p = 0.006) and 0.22 times (p = 0.000) lower. CONCLUSIONS: PJI is the most common indication for secondary TKA revision and within one year after primary TKA. Aseptical failures such as instability, retropatellar arthrosis and aseptical loosening are the predominant reasons for revision more than one year after primary TKA.