9. Neonatal red blood cell transfusion

Despite the frequency of these neonatal RBC transfusions, there is no universally accepted haemoglobin transfusion policy.

Anaemia of prematurity (AOP)

AOP is a multi-factorial condition defined by early (after birth) and significant anaemia that is associated with phlebotomy blood losses, lower erythropoietin (EPO) production and a limited bone marrow response.[2]

Diagnosis of AOP relies upon a combination of parameters such as non-specific clinical symptoms of anaemia as well as haemoglobin and haematocrit levels.[3, 4] However, the exact threshold for haemoglobin or haematocrit levels where inadequate tissue oxygenation (critical anaemia hypoxaemia) definitively occurs in either term or preterm infants is not determined. Overall, this makes the timing of transfusion a neonate extremely challenging.

Systematic reviews comparing liberal versus restrictive RBC transfusion triggers in neonates

Review 1. Whyte & Kirpalani 2011

Included studies: Comparison of low (restrictive) versus high (liberal) haemoglobin levels, four randomised controlled trials, included VLBW infants only
Primary review outcomes: mortality and/or morbidity (ROP grade 3 or greater, severe adverse findings on cranial ultrasound, chronic lung disease, adverse neurodevelopmental outcome)
Results: No statistically significant difference in combined outcomes of death or serious morbidity at first hospital discharge (typical risk ratio 1.19; 95% CI 0.95-1.19), one trial only reported composite of adverse neurodevelopmental outcomes or death at 18-21 months corrected (no difference between groups)
Conclusions: Use of restrictive compared to liberal transfusion thresholds in VLBW infants reduced transfusion exposure but was not associated with any difference is significant morbidity and/or mortality
Comments: given the uncertainties of these conclusions, it would be prudent to avoid haemoglobin levels below the lower limits tested here

Review 2. Bassler et al 2008

Included studies: Benefit to risk ratio of lower versus higher RBC transfusion thresholds, seven randomised controlled studies, included preterm infants ≤ 32 weeks gestation only
Primary review outcomes: apnoea outcomes, ventilation and pulmonary status, ROP, growth, neurologic outcomes, mortality and/or survival with CLD, severe ROP or brain injury, length of stay, blood culture positive sepsis, bowel perforation, NEC, PDA treatment
Results: Adverse outcomes were only described for individual studies and no attempt was made to combine results
Conclusions: Clinical and methodological heterogeneity between studies prevented the authors from pooling any results beyond describing outcomes from individual studies
Comments: Thorough descriptive summary of the individual study findings only

Review 3. Ibrahim et al 2013

Included studies: Comparison of low (restrictive) versus high (liberal) haemoglobin levels, three randomised controlled studies, included VLBW infants only
Primary review outcomes: brain injury on cranial ultrasound, ROP, CLD, NEC or death
Results: No statistically significant differences found between groups for certain morbidities or mortality
Conclusions: Restrictive transfusion thresholds may be used without an associated increased morbidity and/or mortality
Comments: Findings consistent with earlier publications

Review 4. Venkatesh et al 2012

Included studies: Twenty-seven randomised controlled trial included and grouped into:
Trials comparing RBC transfusion and no transfusion/placebo (n=3); different thresholds for transfusion (n=6); differing doses or administration schedule (n=4) or different types/products of RBCs (n=14). Included neonates (preterm/term) <28 days of age at enrolment.
Primary review outcomes: mortality, neurodevelopmental outcome and chronic lung disease
Results: Trials comparing different transfusion thresholds: no significant differences in mortality (RR 1.22, 95% CI 0.84-1.75) or CLD were found
Conclusions: many trials failed to report on outcomes that would considered of primary importance to clinicians with consistent reporting of adverse events is required

Transfusion thresholds for term infants
There are only few studies that follow up on the effects of RBC transfusions in infants. The TRIPICU study [9] shows no difference in oxygenation markers, duration of ventilation, cardiac dysfunction and length of hospital stay between critically ill infants and children that were transfused either with 70g/L or 95g/L RBC preparates.

Age of blood used in neonatal RBC transfusions
The ARIPI trial [10] found that the use of fresh (5.1±2.0 days) RBCs compared with standard (14.6±8.3 days)RBCs did not improve outcomes in very low birth-weight infants receiving RBC transfusions.

Routine use of furosemide in neonatal RBC transfusions
In an attempt to prevent fluid overload, loop diuretic agent furosemide (0.5-2 mg/kg) is used during transfusions in preterm infants. A recent randomised controlled trial demonstrated minimal clinical benefit of co-administered furosemide on cardiopulmonary variables in preterm infants beyond the first week of life. Furthermore, a small increase (p<0.05) in post-transfusion fraction of inspired oxygen (FiO2) levels was observed in the placebo group when compared to the furosemide group.[11] Therefore, furosemide can be beneficial for the cardiopulmonary variables, however its established potential adverse effects should be avoided.

Use of erythropoietin (EPO) in RBC transfusions
Low levels of EPO are frequently used to stimulate RBC production, to avoid RBC drops and to replace RBC transfusions after birth. However, recent studies demonstrate that the use of EPO either early [12] or late [13] in the neonatal course has not been associated with reduced rates of mortality or significant morbidities. Moreover, there is increased risk for retinopathy of prematurity (ROP) with EPO treatment remains a major concern. Darbepoetin is a synthetic form of EPO that may have neuroprotective effects [14] and therefore might be a valueable alternative of EPO.[12]

There is insufficient evidence to either support or disprove a restrictive or liberal RBC transfusion strategy in preterm or term neonates.

However, having a local blood transfusion policy and auditing practice has been shown to reduce the number of neonatal RBC transfusions without any associated increase in adverse outcomes.[15]

When to transfuse?

Whilst there is no evidence determining when to transfuse preterm and term infants, some degree of guidance is required to assist healthcare professionals. Below are suggested pragmatic transfusion thresholds for preterm and term neonates (Table).

Haemoglobin thresholds for transfusion in preterm and term neonates

Safety considerations in neonatal and paediatric transfusion practice

The Serious Hazards of Transfusion (SHOT) reporting scheme has shown that there was a disproportionate number of transfusion errors in the paediatric age group [16]. Besides, the National Comparative Audit of the use of Red Cells in Neonates and Children reported a significant proportion of transfusions prescribed as units but transfused with volumes greater than 20 mL/kg [17]. Furthermore, there are often confusions about maternal and baby samples, multiple births (especially using consecutive identification numbers), babies without first names, failure to apply wristbands, removal of wristbands by children and/or parents or during procedures and failure to make wristbands accessible during surgery (note alternatives may be used).[18] For these reasons, collection of a second sample for confirmation of the ABO group prior to cross-matching is recommended unless reliable electronic patient identification systems are available. In order to reduce neonatal blood testing it may be acceptable to use a cord sample. To prevent over-transfusion of blood components, all prescriptions should be ordered and expressed in millilitres (mL) rather than units. Paedipacks are aliquots made out of one batch of adult blood donation. By using paedipacks, exposure of infant recipients to multiple donors can be avoided. Hospital transfusion laboratories should liaise with neonatal units to develop policies and procedures in order to use paedipacks.

Fetal and neonatal ABO grouping and antibody screening differs from adult grouping. Antibody screening of a fetus/neonate represents the maternal antibody status rather than the fetal/neonatal antibody status. Therefore, within the first four months after birth, samples from both mother and infant should be obtained for initial ABO and D group determination. Besides, antibody screen should also be initiated on the maternal sample when available. Since sample collection from the infant exacerbates the AOP and because it is more convenient to collect sufficient volumes from the mother, a maternal sample is preferred for antibodytesting.

In summary, it is essential to maintain the neonatal and maternal transfusion history (including fetal transfusions) for all new neonatal admissions. Also, it is recommended to obtain a maternal sample for initial testing and for cross-matching when neccessary. Laboratory control measures are required to ensure that donor blood units are ABO and D compatible with both mother and baby; and to determine that it is antigen-negative for clinically significant maternal antibodies.