Coagulase-negative staphylococci are frequently isolated from blood cultures. As these organisms may occasionally cause serious disease, differentiating bacteremia from contamination is very important but often difficult. Over a 26-month period, of 29,542 blood cultures processed at the University of Michigan Medical Center, 2,875 (9.7%) were positive, and of those, 694 (from 527 patients) grew coagulase-negative staphylococci. Isolates from the 439 patients with only one blood culture positive for coagulase-negative staphy-lococci and those from the 18 patients with two positive cultures 10 days or more apart were deemed contaminants. Review of the records of the remaining 70 patients with multiple isolates indicated that 33 had had an episode of true bacteremia, 29 (87.9%) of which were associated with intravascular catheters or prosthetic valves. Overall, 85% of all coagulase-negative staphylococci isolated during the study period were judged to be contaminants. Seventy-one percent of the blood cultures drawn during the episodes of bacteremia were positive for coagulase-negative staphylococci as opposed to only 34% in the patients with contaminated cultures (p < 0.01). Moreover, coagulase-negative staphylococci grew in both aerobic and anaerobic bottles in 85% of blood culture sets drawn during episodes of bacteremia, but in only 30% of the cultures thought to be contaminated (p < 0.001). Growth of coagulase-negative staphylococci in less than 48 hours was also significantly associated with bacteremia (p < 0.01). Antibiotic sensitivity patterns were not useful in differentiating bacteremia from contamination. Thus, clinicians should consider coagulase-negative staphylococci as true blood pathogens in patients with intravascular devices who have a high proportion of blood cultures positive for this organism over a short period of time, and whose cultures became positive in less than 48 hours, with a high percent positive in both bottles. Microbiology laboratories can conserve considerable resources by limiting sensitivity studies to isolates from such patients.

During a 4-month period, 693 patients undergoing surgical procedures were prospectively studied to investigate the etiology of postoperative fever (≥38°C during 48 hours or more). The overall rate of fever was similar for the three categories of surgical procedures studied (14%, 13.4% and 13.1% respectively after clean, clean contaminated and contaminated surgical procedures). No cause of fever was found in 5%, 2.7% and 1.7% of patients who underwent clean, clean contaminated and contaminated surgical procedures. Several episodes of fever were observed for 12 patients after clean surgery; for 11 of them this was due to infection. The mean interval between febrile episodes was 4.7 days. After clean wound surgery, fever documented as infectious began significantly later (2.7 vs 1.6 days) and lasted significantly longer (5.4 vs 3.5 days) than fever for which no source was determined. Only half of the infections were associated with fever.