Cytomegalovirus infection in a neonatal intensive care unit. Blood transfusion practices and incidence of infection

M. P. Griffin, M. O'Shea, J. E. Brazy, J. Koepke, D. Klein, C. Malloy and C. M. Wilfert
Department of Pediatrics, University of Texas Medical Branch, Galveston 77550.

We studied blood transfusion variables and cytomegalovirus (CMV) infection in 385 infants admitted to the Duke University Medical Center, Durham, NC, neonatal intensive care unit over 14 months. Cytomegalovirus antibody titers were measured at birth and monthly thereafter. Urine cultures for CMV were performed regularly. Infants admitted in the first six months (n = 197) received conventionally prepared blood. Infants admitted in the remaining eight months (n = 188) were given frozen, deglycerolized blood. Of the 105 infants weighing 1250 g or less (low birth weight [LBW]), 90 (86%) received transfusions. Two hundred eighty infants weighed more than 1250 g (non-LBW), and 111 (40%) of these were given blood. In the first six months of the study, three infants had CMV viruria. One case was congenital; two were acquired. Both infants who acquired infection were antibody-positive at birth and received multiple transfusions. In the remaining eight months, five infants had CMV viruria. Two cases were congenital; three were acquired. The three infants who acquired infection were antibody-positive at birth and received multiple transfusions. Our study demonstrates that infants with an LBW are more likely to receive blood transfusion and to be given significantly more blood than non-LBW infants. There was no difference in the number of infants acquiring CMV in the two periods despite the use of different preparations of blood.