Pitfalls in the interpretation of leukocyte counts of newborn infants.

Some pitfalls in the interpretation of neonatal leukocyte counts are identified. The well-known variability in neonatal leukocyte counts was investigated by simultaneously sampling arterial, venous and capillary blood, and during periods of rest and mild and violent exercise. Venous blood leukocyte counts were 82% +/- 3.5 (mean +/- SE, P = less than .001) of counts in simultaneously drawn capillary blood from heel punctures; arterial blood counts were 77% +/- 5.3 (P less than .001) of capillary blood values. Following violent crying, capillary blood leukocyte counts increased to 146% +/- 6.1 (P less than .001) of baseline values, and a shift to the left occurred. Milder exercise induced an increase to 113% +/- 5.2 (P less than .05), without a leftward shift. Thus, counts from different vascular sources cannot be considered equivalent. Also, counts from vigorously crying babies may show leukocytosis and a leftward shift, and erroneously suggest bacterial infection. It is recommended that serial counts be obtained from a consistent vascular source in resting babies.