Reassessment of the incubation time in a controlled clinical comparison of the BacT/Alert aerobic FAN bottle and standard anaerobic bottle used aerobically for the detection of bloodstream infections.

This study assessed the minimum incubation time required to detect bloodstream infections during a controlled clinical comparison of the performance characteristics of the BacT/Alert aerobic FAN bottle and the standard anaerobic bottle used aerobically except on a selective basis. Blood was collected from adults with suspected bloodstream infections and inoculated into each bottle, which was monitored in the BacT/Alert Microbial Detection System. The anaerobic bottle was vented before incubation except when cultures were obtained from patients on the colorectal and gynecologic surgical and emergency services. Statistical analysis was limited to those culture sets in which each bottle was inoculated with > or = 8 mL of blood and bacterial growth was considered to be clinically significant. A total of 682 positive cultures from 243 patients satisfied the inclusion criteria. Significantly more isolates of Staphylococcus aureus (p < 0.001), S. epidermidis (p < 0.001), other coagulase-negative staphylococci (p < 0.001), Enterococcus spp. (p = 0.04), Escherichia coli (p = 0.03), all Enterobacteriaceae (p < 0.001), Pseudomonas aeruginosa (p = 0.001), and Candida spp. (p < 0.001) were detected by the aerobic FAN bottle. Significantly more septic episodes due to S. aureus, S. epidermidis, other coagulase-negative staphylococci, Enterobacteriaceae, P. aeruginosa, and Candida spp. were detected by the aerobic FAN bottle. Significantly more bacterial isolates were detected by the aerobic FAN whether or not antibiotics were being administered at the time of blood culture, whereas there were significantly fewer positive cultures in the vented standard anaerobic bottle when patients were receiving antimicrobial therapy than when they were not. All but 5% of positive cultures were detected within three days. Only six of the cultures requiring four or five days of incubation represented true misses, and only one of these six resulted in a change in therapy which, however, did not affect the patent's outcome.