Objective: To determine whether the immature neutrophil (band) count in the peripheral blood smear helps to distinguish young febrile children with bacterial or respiratory viral infections.
Design and setting: A prospective cohort study in 3 pediatric emergency departments.
Patients: A convenience sample of 100 febrile children aged 2 years or younger with either laboratory-documented bacterial infections (n = 31; 24 with urinary tract infections, 7 with bacteremia) or laboratory-documented respiratory viral infections (n = 69). Each patient received a clinical appearance score using the Yale Observation Scale prior to laboratory evaluation. A complete blood cell count was obtained from all patients and manual differential count of the peripheral blood smear was performed by 1 senior technician masked to clinical information.
Main outcome measure: Band counts, represented as a percentage of white blood cells in the peripheral blood smear, the absolute band count, and band-neutrophil ratio. Logistic regression analysis was performed to determine whether the band count helps to distinguish bacterial infections from viral infections after adjusting for age, temperature, Yale Observation Scale score, and absolute neutrophil count.
Results: Patients with bacterial infections had a higher mean absolute neutrophil count (11.3 vs 5.9 x 10(9)/L; P<.01) than patients with respiratory viral infections. There was no difference, however, in percentage band count (13.5% vs 13.3%; P = .90), absolute band count (2.2 vs 1.9 X 10(9)/L; P= .31), or band-neutrophil ratio (0.24 vs 0.33; P = .08, bacterial vs viral, respectively); the band count did not help to distinguish bacterial and viral infections after adjusting for age, temperature, Yale Observation Scale score, and absolute neutrophil count in the regression analysis.
Conclusion: The band count in the peripheral blood smear does not routinely help to distinguish bacterial infections from respiratory viral infections in young febrile children.