Neonatal sepsis resulting from possible amniotic fluid infection: risk and detection.

Antibiotics are frequently prescribed "prophylactically" when the neonate is considered to be at risk for infection. The risk factors of prolonged rupture of membranes (greater than 24 hours), maternal fever/infection, and/or unexplained preterm labor (suggesting possible anmiotic fluid infection) were investigated in 276 babies. Only two of 150 babies investigated for a single factor proved to have sepsis, while of the 126 babies who had multiple factors, 13 had sepsis. Several laboratory tests, used singly or in combination, were more helpful than clinical manifestations in predicting which babies were likely to be infected. Neonatal sepsis was present in 6 per cent of the total, in 10 per cent of neonates with clinical signs, in 21 per cent with an increased immature/total neutrophil ratio (I/T ratio greater than or equal to 0.2), and in 36 per cent of infants with a positive "sepsis screen." The incidence of "infection" (sepsis and "very probable" infection) was 12 per cent overall, but was 14 per cent with neonatal signs, 44 per cent with I/T ratio greater than or equal to 0.2, and 74 per cent in those with a positive sepsis screen. A leukocyte count less than 5,000/cu mm and/or an I/T ratio greater than or equal to 0.2 was 100 per cent sensitive for sepsis, but the sepsis screen was most "efficient" at detecting "infection." Starting antibiotics on the basis of risk factors alone does not seem appropriate. In situations where amniotic fluid infection is possible, evaluation with the leukocyte count and differential (with or without other tests) could decrease the indiscriminate use of antibiotics, particularly when a single risk factor is the reason for suspecting infection.