Eosinophilia in the hospitalized neonate.

Abstract

The incidence of and clinical associations with eosinophilia in a cross-section of hospitalized newborns had not been studied previously. The medical records of 200 such infants less than 29 days of age were reviewed. Total eosinophil counts were calculated and associations with a variety of clinical factors were examined. Eosinophilia (adjusted for day of age) occurred at least once in 22% of the infants studied. No significant skew could be detected in the age of onset of eosinophilia. Statistical analysis revealed eosinophilia to be significantly associated (p less than .05) with length of hospitalization, number of days of antibiotic therapy and use of parenteral alimentation. A trend toward eosinophilia was noted in lower birth weight infants (0.05 less than p less than 0.10). Eosinophilia was not associated with sex, age, Apgar score at birth, Caesarian section delivery, transfusions, phototherapy, specific diagnoses or type of oral feeding. Multiple regression analysis showed a highly positive correlation (r = .81) of associated factors with hospitalized days as the dependent variable. Eosinophilia appears to be a non-specific finding in sick neonates and is related to the severity of illness and the number of diagnostic or therapeutic procedures performed.