The purpose of this study was to determine an association between packed red blood cell (PRBC) transfusions for anemia and necrotizing enterocolitis (NEC) in a subset of stable, growing, premature neonates. As part of a survey of current clinical practices over a 17-month period from June 1999 to October 2000, a chart review was performed to determine the relationship between elective PRBC transfusions and the occurrence of NEC. Demographic data were tabulated and compared between the NEC patients with a prior history of immediate blood transfusion (within 48 hours of onset of symptoms) and those NEC patients without a prior history of immediate blood transfusion. A total of 908 (inborn) neonatal admissions had received 751 PRBC transfusions during the study period; of these, 17 patients (1.8%) had developed radiographic, clinical, or surgical signs of NEC. Six cases of NEC (35%; six of 17 patients) were associated with PRBC transfusions (0.8%; six of 751 transfusions). The transfusion-associated NEC group developed presenting signs within 22 +/- 5 hours (median, 19; range, 12 to 38) of a PRBC transfusion at a mean age of 32 +/- 7 days. In contrast, the non-transfusion-associated NEC group (n = 11) had onset of NEC at a mean age of 12 +/- 7 days ( P < 0.05) after 185 +/- 91 hours (median, 180; range, 96 to 312; P < 0.02] of a transfusion. Prior to the onset of NEC, all of the neonates in the transfusion-associated NEC group were stable, growing, not ventilated, receiving full enteral feedings, and had no other active medical problems except anemia (hematocrit, 24 +/- 3%). In contrast, the nontransfusion NEC group was more often ventilated, was receiving < 50% of fluids by mouth, had lower Apgar scores, and was transfused for an average hematocrit of 37 +/- 7% ( P < 0.05). There was no significant difference in the type, storage, volume, or preservative used between the blood products in the two groups. We identified an unanticipated relationship between late-onset NEC in stable, growing, premature neonates who were transfused electively for anemia of prematurity.

Preterm infants often receive multiple packed red blood cell (PRBC) transfusions that are intended to improve tissue oxygen levels. Near-infrared spectroscopy (NIRS) monitors regional cerebral tissue oxygen saturation (CrSO(2)) and splanchnic tissue oxygen saturation (SrSO(2)). Before such technology can be employed in neonatal transfusion management, it must first be established that transfusions result in an increase in tissue oxygen saturation. This prospective, observational study used NIRS to determine if PRBC transfusions increase the CrSO(2) and SrSO(2) of symptomatic anemic premature neonates. CrSO(2) and SrSO(2) values were compared for 20-minute duration immediately before, during, immediately after, and 12 hours after transfusion. As a secondary objective, CrSO(2) and SrSO(2) values were correlated with hemoglobin (Hgb) levels. One-way analysis of variance and Pearson correlation statistical tests were used for analysis. A statistically significant increase in CrSO(2) and SrSO(2) values were observed after transfusion in the 30 subjects included (CrSO(2): 62.8 +/- 1.6, 65.6 +/- 1.7, 68.0 +/- 1.3, 67.6 +/- 1.4, P < 0.001 and SrSO(2): 41.3 +/- 2.2, 46.7 +/- 3.0, 52.1 +/- 2.8, 48.2 +/- 2.5, P < 0.001). No correlation was found between CrSO(2) or SrSO(2) and Hgb values. NIRS identified increases in CrSO(2) and SrSO(2) in preterm neonates after PRBC transfusions and has the potential to become incorporated into neonatal transfusion management paradigms.

No abstract