Bacteremia is commonly suspected in sick neonates. BC are considered negative after 3 days while final laboratory report may take 5 days. Reduction of hospital stay by one or more days could have significant financial and psychosocial implications. We undertook this retrospective study to test the hypothesis that 48 hours are sufficient to declare a BC negative. We reviewed all BC records of newborns admitted to our tertiary Neonatal Intensive Care Unit (NICU) between 01/01/1992 and 12/31/1994. Data collected included patient's age at the time of sampling, incubation period of BC, identification and antimicrobial susceptibility of organisms. Positive BC were classified into perinatally-acquired or nosocomial if BC was obtained before or after 3 days of age, respectively. BC were obtained from 740 patients, 184 (24.8%) of whom had positive BC. Perinatal bacteremia was present in 15 (8.1%) and was detected at 1.6 ± 0.52 (mean ± SD) days (none after 48 hours). Nosocomial bacteremia was detected after 2.32 ± 0.7 days (range 1-6 days). Escherichia coli and Group B streptococci accounted for all perinatal infections, while coagulase negative staphylococci,Staphylococcus aureus, Klebsiella and Enterobacter were the predominant organisms in nosocomial infections. We conclude that perinatally-acquired bacteremia can be excluded after two days of incubation. The majority of neonatal BC are negative. In babies in whom treatment will not be continued if cultures are negative, the cost savings of a length of stay shortened by one day might be considerable. Nosocomial bacteremia in NICU patients is usually confirmed within 2 to 3 days, but BC may take up to 6 days to yield a pathogen such as Enterobacter or Moraxella.
