In a controlled evaluation of 6,010 blood cultures, the yield of clinically significant microorganisms was greater from a lysis-centrifugation system (Isolator, Du Pont Co.) than from a nonvented vacuum bottle containing tryptic soy broth with sodium polyanetholesulfonate and CO2 and a vented bottle containing biphasic brain heart infusion medium with sodium polyanetholesulfonate. The Isolator significantly increased the frequency of isolation of Staphylococcus aureus and Candida spp. and significantly decreased the time required for the detection of S. aureus, Pseudomonas aeruginosa, and Candida spp.; however, anaerobic bacteria were recovered significantly more frequently from nonvented bottles with tryptic soy broth, and pneumococci were recovered significantly more frequently from both bottle systems. Contamination of cultures was significantly greater with the Isolator system than with either bottle system. Regardless of the number of blood cultures obtained per septic episode, the Isolator detected microbiologically proven bacteremia or fungemia in a significantly greater number of patients and significantly decreased the time required for detection.

A parallel study of blood cultured conventionally and following processing in an Antimicrobial Removal Device (ARD) was conducted with patients suspected of being clinically bacteremic. Cultures yielded 205 isolates from 87 clinically bacteremic patients, of whom 28 were receiving antibiotics at the time their cultures were performed. Overall, ARD processing neither increased the isolation rate nor decreased the time interval required for detection of organisms compared with conventional processing. ARD processing was the only means of isolation of organisms from only seven of the 87 patients and three of the 28 receiving antibiotics, whereas conventional processing accounted for the only means of isolation of organisms from 17 of the 87 patients and nine of the 28 receiving antibiotics. ARD processing demonstrated no advantages over conventional processing of blood cultures.

In a comparison of 1,368 positive blood cultures, a vented Roche Septi-Chek (V-RSC) blood culture bottle was superior to an unvented tryptic soy broth-containing bottle (Difco) for the recovery of all aerobic and facultatively anaerobic microorganisms. Anaerobic bacteria were recovered more frequently and earlier in the unvented tryptic soy broth-containing bottle. A separate comparison of 529 positive blood cultures was conducted to examine the performance of the V-RSC bottle with that of a vented brain heart infusion biphasic medium. The V-RSC bottle recovered significantly more isolates of Enterobacteriaceae and of anaerobic bacteria than did the vented brain heart infusion biphasic medium. The V-RSC bottle is a reliable blood culture system for all aerobic and facultatively anaerobic microorganisms. Because of its suboptimal recovery of anaerobic bacteria, it is recommended that the V-RSC bottle be used in combination with an unvented vacuum blood culture bottle.

No abstract