PURPOSE OF THE STUDY: To analyse risk factors for the development of prosthetic infection after total joint arthroplasty in our group of patients, and to contribute to the understanding of conditions leading to infection and, consequently, to its eliminations. MATERIAL: We assessed a group of 196 patients who were treated for deep infection of total hip or knee arthroplasty at our department between 1990 and 2006. The group was compared with a group of 400 randomly selected patients who, in the same period, underwent total hip or knee arthroplasty and were free from infectious complications. METHODS: Several factors considered to be associated with a potential risk for the development of prosthetic infection were selected and the frequency of their occurrence in the two groups (infected and non-infected) was assessed. When the rate o its occurrence in the infected group was higher than five-times that in the non-infected group, the factor was regarded as significant. When its occurrence was less frequent, the factor was considered less important for the development of prosthesis infection. RESULTS: The following factors were shown to be significant (the figure in brackets denotes the difference in frequency between the infected and the non-infected group): * Body mass index lower than 20 (6.5) or higher than 60 (5.1) * Diabetes mellitus on insulin therapy (5.3) * Chronic renal insufficiency in the patient's medical history (5.6) * Urinary tract infection in the patient's history (7.8) * Immunological deficiency in the patient's history (6.1) * Alcohol abuse (8.2) and serious psychiatric disease (5.1) in the patient's history * Surgery involving post-traumatic or traumatic tissues (6.3) * Presence of erysipelas (28.6) and varicose ulcer (8.8) on the leg operated on * Presence of osteomyelitis and articular infection in the region of implantation In each patient, accumulation of factors was assessed and it appeared that when any three (or more) factors investigated were found in a patient, the probability of prosthesis infection increased 16-times; when two (or more) significant factors were present, the probability of infection development was 14-times higher. DISCUSSION: We compared these results with the relevant literature data, and the majority of them were in agreement with the other authors. In contrast to the published information, we found a higher occurrence of psychiatric disorders in the infected group. Of all the factors investigated, the presence of erysipelas appeared to be most significant, with a frequency 28.6-fold higher in the infected than the non-infected group. References to this risk factor were rare in the literature. CONCLUSIONS: We recommend a thorough evaluation of risk factors in the patients planned to undergo total arthroplasty, assessment of factor significance and accumulation, and careful consideration of all pros and cons in patients at high risk of infection development.

1 ortopedická klinika MU a FN u sv Anny v Brne

Prevention of Prosthetic Joint Infection: From Traditional Approaches towards Quality Improvement and Data Mining