Factors influencing the incidence and outcome of infection following total joint arthroplasty.

During a ten-year period, 4240 total hip, knee, and elbow arthroplasties were performed. The overall infection rate was 1.25%. Certain groups were identified as being at higher risk of infection following total joint arthroplasty: rheumatoid arthritics were at 2.6 times greater risk than osteoarthritics; patients undergoing total hip arthroplasty as a revision of a previous operation were eight times more likely to have infection than those undergoing a primary operation; and patients with metal-to-metal hinged knee prostheses, when compared with patients with metal-to-plastic knee prostheses, were 20 times more likely to have infection. The majority of infections could be attributed either to perioperative problems or late bacterial seeding from a distant site. Although most infections occurred by two years after operation, late infections, particularly in rheumatoid patients via the hematogenous route, occurred as long as nine years after operation. There was no correlation between the Gram's-staining characteristics of the pathogen and the outcome of the infected joint. Gram-negative organisms were frequent in the perioperative period and reflected either nosocomial infection or the ineffectiveness of the prophylactic antibiotic regimen used in inhibiting gram-negative pathogens. The major factors that influenced the outcome of the infected joint included the interval from the initial surgery to recognition of infection, the delay in institution of appropriate treatment, the particular joint that was infected, the integrity of the bone-cement interface, the type of prosthesis used, and the host susceptibility. Identification of high-risk groups and the recognition that patients with joint implants are at risk of infection at any time in the postoperative period may lead to a lowered infection rate in the future.