Do cardiac output and serum lactate levels indicate blood transfusion requirements in anemia of prematurity?

BACKGROUND: Whether and when to transfuse in anemia of prematurity is highly controversial. Some authors suggest transfusions simply if the hemoglobin (Hb) level is below a defined normal range. Others propose the use of clinical or laboratory parameters in anemic patients to decide whether to transfuse or not.

HYPOTHESIS: A decreasing amount of circulating Hb should cause a compensatory increase in cardiac output (CO) and an increase in arterial serum lactate.

MATERIALS AND METHODS: In 56 anemic preterm infants (not in respiratory or hemodynamic failure) we analyzed CO after the first week of life using a Doppler sonographic method. At the same time serum lactate levels, Hb levels and oxygen saturation were registered. Nineteen of these patients were given transfusion when they demonstrated clinical signs of anemia by tachycardia > 180/min, tachypnea, retractions, apneas and centralization (group 2). The remaining 37 patients were not transfused (group 1). Serum lactate, CO, heart rate (HR), oxygen delivery, respiratory rate, capillary refill and Hb were analyzed in both groups and in group 2 before and 12-24 h after transfusion. Data between groups 1 and 2 and in group 2 before and after transfusion were compared.

RESULTS: In the 56 patients studied no linear correlation between Hb and CO or between Hb and serum lactate was found. Nor could any correlation be demonstrated between the other variables studied. Examining the subgroups separately, a negative linear correlation was demonstrated between serum lactate and oxygen delivery in group 2. No other significant correlations were detected. However, when the pre- and post-transfusion data were compared in group 2 (increase of Hb from 9.45 (SD 3.44) to 12.5 (SD 3.8) g/100 ml), the CO decreased from 281.3 (SD 162.6) to 224 (SD 95.7) ml/kg per min (p < 0.01) and serum lactate decreased significantly from 3.23 mmol/l (SD 2.07) before to 1.71 (SD 0.83) after transfusion. Oxygen delivery was 35.8 (+/- 0.19) ml/kg per min group 1, 27.8 (+/- 0.05) pre- and 43.4 (+/- 0.07) post-transfusion in group 2 (p < 0.01).

CONCLUSIONS: CO measurements and serum lactate levels add little information to the decision-making process for blood transfusions, as neither CO nor serum lactate levels correlate with HB levels in an otherwise asymptomatic population of preterm infants. In infants where the indication for blood transfusion is made based on traditionally accepted clinical criteria, serum lactate is an additional laboratory indicator of impaired oxygenation, as it correlates significantly with oxygen delivery. A significant lower oxygen delivery in patients in whom blood transfusion is indicated and an increase in oxygen induced by transfusion demonstrate the value of these criteria in identifying preterm infants who benefit from transfusion.