Abstract
Aims
The microbiological detection of microorganisms plays a crucial role in the diagnosis as well as in the targeted systemic and local antibiotic therapy of periprosthetic infections (PJI). Despite extensive efforts to improve the sensitivity of current culture methods, the rate of culture-negative infections is approximately 10–20% of all PJI. This study investigates an preanalytical algorithm (culture collection and direct processing in the OR) to potentially increasing culture yield in patients with PJI.
Methods
Patients undergoing staged revision arthroplasty for PJI in our hospital between October 2021 and 2022 were included in this prospective pilot study. Intraoperatively twenty tissue samples were collected and distributed among 4 groups. Tissue samples were prepared according to standard without medium and in thioglycolate medium at 3 different temperatures (room temperature, 4°C, 37° for 24h before transport to microbiology) directly in the OR. The removed implants were sonicated. Cultures were investigated on days 1, 3, 7, 12, 14 for possible growth. All grown organism, the number of positive samples and the time to positivity were recorded and compared.
Results
71 patients were included (age, gender). Compared to the standard procedure the thioglycolate broth at 37°C was significantly more often culture-negative (p=0.031). No significant differences in the frequency of culture-negative samples were detected in the other groups.
8.4% (6/71) patients were culture negative in the standard culture but positive in the thioglycolate samples. In contrast, 7% (5/71) were culture negative in the thioglycolate samples but had bacterial detection in the standard approach. In 4.7% (3/63) of the patients, only the sonication showed growth, whereas 25.4% (16/63) had no growth in sonication fluid but in one of the cultures.
For S. caprae, there was a significantly different distribution (p=0.026) with more frequent detection in the group with thioglycolate at 37°C.
The standard procedure (p=0.005) and sonication (p=0.023) showed a shorter time to positivity of the culture compared to the thioglycolate approach at 4°C.
Conclusions
No general differences could be shown between the standard preparation and the thioglycolate preparation; in particular, storage at different temperatures does not seem to result in any difference. For individual cases (8% in this study), bacterial growth was detected in the thioglycolate group that would have been culture-negative otherwise. There might be organism dependent differences in growth in different media.
