Objective: To define a subset of very low birth weight (VLBW) infants who might benefit from recombinant human erythropoietin (r-HuEPO) treatment.
Study design: We reviewed the records for all VLBW (birth weight (BW) < or = 1500 gm) infants who were admitted to our nursery within the first 3 days of life between January 1991 and December 1994 and discharged alive.
Results: These infants received an average of 2.02 transfusions, far fewer than the 7 to 11 previously reported for VLBW infants. Infants with a BW of 1251 to 1500 gm received very few transfusions. More than three quarters of transfused infants received a transfusion in the first 2 weeks of life before r-HuEPO would be expected to be effective. Assigning units to individual infants and holding the units for 14 days, a practice adopted in our blood bank in 1993, resulted in a 44% decrease in donor exposures in infants receiving more than one transfusion. Holding assigned units for 30 days, a practice our blood bank has now adopted, should result in 56% of all transfused infants having a single donor exposure and 89% having one or two donor exposures. Cost-benefit analysis only supports routine use of r-HuEPO in infants weighing less than 750 gm.
Conclusion: VLBW infants receive far fewer transfusions than the number previously reported. Assigning units to individual patients and holding those units for 30 days, together with efforts to minimize the need for transfusions make routine use of r-HuEPO unnecessary.