Background: The value of microbiological culture to diagnose late periprosthetic infection is limited, especially because standard methods may fail to detect biofilm-forming sessile or other fastidious bacteria. There is no agreement on the appropriate cultivation period, although this period is a crucial factor. This study was designed to assess the duration of culture that is necessary for reliable detection.
Patients and methods: Ten periprosthetic tissue specimens each were obtained during revision from 284 patients with suspected late hip or knee arthroplasty infection. Five samples were examined by microbiological culture over a 14-day period, and 5 were subjected to histologic analysis. To define infection, a pre-established algorithm was used; this included detection of indistinguishable organisms in >/=2 tissue samples or growth in 1 tissue sample and a positive result of histologic analysis (>5 neutrophils in at least 10 high-power fields). The time to detection of organisms was monitored.
Results: Infection was diagnosed in 110 patients. After 7 days (the longest incubation period most frequently reported), the detection rate via culture was merely 73.6%. Organisms indicating infection were found for up to 13 days. "Early"-detected species (mostly staphylococci) emerged predominantly during the first week, whereas "late"-detected agents (mostly Propionibacterium species) were detected mainly during the second week. In both populations, an unequivocal correlation between the number of culture-positive tissue samples and positive results of histologic analysis was noted, which corroborated the evidence that true infections were detected over the entire cultivation period.
Conclusions: Prolonged microbiological culture for 2 weeks is promising because it yields signs of periprosthetic infection in a significant proportion of patients that would otherwise remain unidentified.