Detection of neonatal sepsis of late onset

JAMA. 1982 Jan 22-29;247(4):489-92.


Five tests have been used in combination to diagnose neonatal sepsis. This study describes their use in 56 infants evaluated between 8 and 60 days of age, who had nonspecific signs of infection at presentation, as well as further evidence in 524 infants evaluated in the first week after birth. When two or more of the five tests had abnormal results (leukocyte count less than 5,000/ cu mm; immature/total neutrophils greater than or equal to 0.2; ESR greater than or equal to 15 mm/hr; latex C-reactive protein, positive; and latex haptoglobin, positive), a "sepsis screen" was considered positive. A positive screen was found in 23 infants, ten of whom had proved sepsis, and only two had no evidence of infection. With the addition of a leukocyte count greater than 20,000/cu mm, the remaining two cases of sepsis would have been detected. In those with a negative screen (n = 33), 26 had no evidence of infection. The sepsis screen seems to be a useful adjunct in the diagnosis of neonatal sepsis during and beyond the first week.

MeSH terms

  • Bacterial Infections / diagnosis
  • Birth Weight
  • Female
  • Follow-Up Studies
  • Humans
  • Infant
  • Infant, Newborn
  • Infant, Newborn, Diseases / diagnosis*
  • Infant, Premature, Diseases / diagnosis
  • Infections / diagnosis*
  • Male
  • Methods
  • Time Factors