Risk factors associated with deep surgical site infections after primary total knee arthroplasty: an analysis of 56,216 knees.

BACKGROUND Deep surgical site infection following total knee arthroplasty is a devastating complication. Patient and surgical risk factors for this complication have not been thoroughly examined. The purpose of this study was to evaluate risk factors associated with deep surgical site infection following total knee arthroplasty in a large U.S. integrated health-care system. METHODS A retrospective review of a prospectively followed cohort of primary total knee arthroplasties recorded in a total joint replacement registry from 2001 to 2009 was conducted. Records were screened for deep surgical site infection with use of a validated algorithm, and the results were adjudicated by chart review. Patient factors, surgical factors, and surgeon and hospital characteristics were identified with use of the total joint replacement registry. Cox regression models were used to assess risk factors associated with deep surgical site infection. RESULTS A total of 56,216 total knee arthroplasties were identified; 63.0% were done in women, the average age of the patients was 67.4 years (standard deviation [SD] = 9.6), and the average body mass index (BMI) was 32 kg/m2 (SD = 6). The incidence of deep surgical site infection was 0.72% (404/56,216). In a fully adjusted model, patient factors associated with deep surgical site infection included a BMI of ≥35 (hazard ratio [HR] = 1.47), diabetes mellitus (HR = 1.28), male sex (HR = 1.89), an American Society of Anesthesiologists (ASA) score of ≥3 (HR = 1.65), a diagnosis of osteonecrosis (HR = 3.65), and a diagnosis of posttraumatic arthritis (HR = 3.23). Hispanic race was protective (HR = 0.69). Protective surgical factors included use of antibiotic irrigation (HR = 0.67), a bilateral procedure (HR = 0.51), and a lower annual hospital volume (HR = 0.33). Surgical risk factors included quadriceps-release exposure (HR = 4.76) and the use of antibiotic-laden cement (HR = 1.53). In a subanalysis, operative time was a risk factor, with a 9% increased risk per fifteen-minute increment. CONCLUSIONS Use of a comprehensive infection surveillance system, combined with a total joint replacement registry, identified patient and surgical factors associated with infection following total knee arthroplasty in a large sample. High-risk patients should be counseled, and modifiable clinical conditions should be optimized. Use of antibiotic irrigation should be encouraged, but antibiotic-laden cement may not be useful. LEVEL OF EVIDENCE Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.