Clinical impact of bacteria and fungi recovered only from broth cultures

We prospectively evaluated 356 bacteria and fungi recovered from broth enrichment tubes from cultures with sterile direct plates to determine the clinical impact of isolates recovered only from broth cultures. These "broth only" isolates (BOI) were classified as contaminants or true on the basis of review of patient charts. True isolates were considered clinically relevant only if they altered or should have altered patient management. Of 356 BOI, 259 (73%) were considered contaminants (mostly coagulase-negative staphylococci and Propionibacterium spp.) and 97 (27%) were considered true. For individual microorganisms, 9 of 9 (100%) Staphylococcus aureus isolates, 13 of 13 (100%) members of the family Enterobacteriaceae, 10 of 12 (83%) fungi, 7 of 10 (70%) enterococci, 7 of 11 (64%) other gram-negative bacilli, 13 of 31 (45%) anaerobic bacteria, 10 of 24 (42%) streptococci, 22 of 140 (16%) coagulase-negative staphylococci, 6 of 92 (7%) Propionibacterium spp., and 0 of 14 (0%) diphtheroids and Bacillus spp. were classified as true. Eleven of 97 (11%) patients with true BOI had clinically relevant isolates. Fifty-nine of the 97 (61%) patients with true isolates already were on therapy, and no change was made because of the BOI. Six (6%) patients with contaminants received therapy for their BOI. We conclude that broth inoculated as an adjunct to direct plating seldom yields results that favorably alter patient management and could be omitted for most specimens without compromising patient care.