To retrospectively analyze the etiology and microbiological results of synovial fluid culture, periprosthetic tissue culture and sonication fluid culture in 74 episodes of prosthetic joint infection (PJI) in 66 patients treated at the Department for Orthopaedic Surgery in 5 years period.

PJI was diagnosed according to the standard definition criteria (1). Conventional microbiological methods were used together with sonication of explanted prosthesis and sonication fluid culture.

The results are presented in Table 1. Acute PJI were more common in female patients (72%), whereas delayed PJI occurred in both sexes equally. Surprisingly, among acute PJI streptococcal (especially Streptococcus agalactiae) PJI were as common as Staphylococcus aureus PJI; we also observed high percentage of polymicrobial PJI (26 % among acute PJI and 30 % among delayed PJI), whereas 6/21 (28.6%) were detected by sonication fluid culture only. Preoperative synovial fluid culture identified the pathogen in 91.3% of acute PJI and in 63.2% of delayed PJI. Sonication fluid culture identified causative pathogens in 85% of acute PJI and in 95% of delayed PJI. In acute PJI 24/34 (70.6%) patient had concordant microbiological results between standard cultures and sonication fluid cultures, whereas in 4/34 (11.7%) microorganisms were detected by sonication fluid culture only and in 6/34 (17.6%)by tissue culture only. 23/40 (57.5%)patients with delayed PJI had concordant microbiological results; in 3/40 (7.5%) pathogens were detected by sonication fluid culture only.

We found sonication as very useful additional method in diagnostics of PJI, especially low grade PJI. At the same time we observed high percentage of positive and concordant mcrobiological results between all three microbiological methods in patients with acute PJI.

Introduction: Femoral fracture after total or hemiar-throplasty is an uncommon but often difþcult complication to treat. Its incidence varies from 1% after primary THA to 4% after revision THA. The goal of our study was to determine the healing rate of the fractures and the clinical outcome of the patients analysed against type of fractures, method of þxation and loosening of the prostheses. Methods: We retrospectively analysed 63 operatively treated patients with periprosthetic femoral fractures operated on in two large institutions from the year 1994 with at least 12 month of follow-up. Fractures were classiþed according the Vancouver classiþcation system. Perioperative femoral fractures were excluded from the study. Results: According the Beals and Tower we had 38 excellent, 13 good and 7 poor results, one patient died in the early postoperative period and 4 were lost to follow-up. 16 procedures were associated with a perioperative complication. Conclusions: We think that fractures associated with a loose interface cemented or cementless are best treated by removal of the implant and insertion of a long stemmed prosthesis in the reduced femur. A periprosthetic fracture associated with a stable prosthesis can be safely treated by a reduction and internal þxation. The use of cement in revision is a good choise in the older patient who needs a fast mobilisation and early full weight bearing.