Total knee arthroplasty (TKA) has been proven to be the most effective treatment for patients with severe or “end-stage” joint disease. Although infection is not a frequent complication of total knee arthroplasty, it is certainly one of the most dreaded. The purpose of this study was to identify related factors associated with septic arthritis.

2202 primary total knee arthroplasties were done in 1257 patients between 1995 and 2006. Of these knee arthroplasties, 2022 knees in 1146 patients were available for follow-up. Revision arthroplasty procedures and infected knees were excluded. 252 knees in 147 males, 1770 knees in 999 females were done. Their mean age at the time of primary TKA was 70.6 (range, 26–91) years. The mean follow-up period post primary TKA was 48 (range, 3–145) months. The medical records were reviewed to extract the following information: age, gender, body mass index, preoperative CRP, preoperative ESR, preoperative TP, duration of surgery, operative blood loss, total blood loss, duration of surgical drain, duration of antibiotic prophylaxis, primary diagnoses, smoking, diabetes mellitus, steroid or DMARDs therapy, previous operation around the knee joint, previous arthroscopic surgery, previous except arthroscopic surgery, previous operation of high tibial osteotomy (HTO) or open reduction internal fixation (ORIF), residue of internal fixation material, bone graft, patella replacement, and bone cement.

Proportions were compared using the chi-square or two-tailed Fisher’s exact test, as appropriate. Continuous variables were compared by the student’s t-test. Logistic regression analysis (stepwise) of selected variables from univariate analysis was performed to identify factors independently associated with the development of infection following total knee arthroplasty.

During the study period, 17 infected knee arthroplasties in 17 patients were identified. The infections occurred in 8 males and 9 females, with a medial age of 69.5 years.

The results of univariate analysis indicating those variables statistically associated with infection are : gender (p < 0.0001), smoking (p = 0.02), previous operation around the knee joint (p = 0.001), previous except arthroscopic surgery(p < 0.0001), previous operation of ORIF (p < 0.0001), residue of internal fixation material (p < 0.0001).

Logistic regression analysis indicated that the four predictors of infection following total knee arthroplasty were gender (odds ratio [OR], 0.2; 95% confidence interval [CI95], 0.1 to 0.6; P=0.005), previous operation of ORIF (OR, 7.9; CI95, 1.1 to 57.1; P=0.041), residue of internal fixation material (OR, 26.0; CI95, 4.5 to 151.0; P< 0.001), body mass index (OR, 1.2; CI95, 1.0 to 1.3; P=0.007).

We conclude that the risk factors of infection after TKA were previous operation of ORIF, gender, residue of internal fixation, and body mass index.