Impaired sensitivity of a single early leukocyte count in screening for neonatal sepsis.

The common clinical practice of using a single, early white blood cell (WBC) count to screen for early onset neonatal sepsis was investigated in a population of 61 newborn infants with culture proven sepsis in the first 3 days of life. Thirteen patients (21%) had a falsely normal WBC screening test. The patients with true positive and falsely normal WBC counts did not differ by risk factors for sepsis, birth weight, age, outcome or severity of disease. However, there was a significant delay between the screening test and the positive blood culture in the patients with false normal WBC counts and not in the patients with positive abnormal WBC counts (14.9 +/- 5.9 hours vs. 2.8 +/- 1.4 hr, mean +/- SE, P less than 0.001). A WBC count obtained soon after birth as currently utilized may not adequately screen for early onset neonatal sepsis.