Effectiveness of Granulocyte Colony-Stimulating Factor in Hospitalized Infants with Neutropenia

Objective The objective of this study was to determine the time to hematologic recovery and the incidence of secondary sepsis and mortality among neutropenic infants treated or not treated with granulocyte colony-stimulating factor (G-CSF). Study Design We identified all neutropenic infants discharged from 348 neonatal intensive care units from 1997 to 2012. Neutropenia was defined as an absolute neutrophil count <= 1,500/mu L for >= 1 day during the first 120 days of life. Incidence of secondary sepsis and mortality and number of days required to reach an absolute neutrophil count > 1,500/mu L for infants exposed to G-CSF were compared with those of unexposed infants. Results We identified 30,705 neutropenic infants, including 2,142 infants (7%) treated with G-CSF. Treated infants had a shorter adjusted time to hematologic recovery (hazard ratio: 1.36, 95% confidence interval [CI]: 1.30-1.44) and higher adjusted odds of secondary sepsis (odds ratio [OR]: 1.50, 95% CI: 1.20-1.87), death (OR: 1.33, 95% CI: 1.05-1.68), and the combined outcome of sepsis or death (OR: 1.41, 95% CI: 1.19-1.67) at day 14 compared with untreated infants. These differences persisted at day 28. Conclusion G-CSF treatment decreased the time to hematologic recovery but was associated with increased odds of secondary sepsis and mortality in neutropenic infants. G-CSF should not routinely be used for infants with neutropenia.