Premature infants cared for in Australian neonatal intensive care units lose blood at a rate that would be alarming in older patients, yet their physiology tolerates low haemoglobin values far better than adults. Neonatologists in Sydney, Melbourne, Brisbane, Perth and Adelaide balance this paradox every shift, weighing the cumulative phlebotomy losses of a 26-week infant against a developing marrow that cannot yet mount an adequate erythropoietic response. The two therapeutic levers clinicians use are recombinant erythropoietin and packed red cell transfusion, and the decision of when to use each shapes short-term stability as well as long-term neurodevelopmental outcomes.
Australian practice draws on data from the Australian and New Zealand Neonatal Network, local transfusion guidance issued through the National Blood Authority, and shared learning with perinatal societies across the region. Resources curated through platforms such as the FAOPS 2020 conference site continue to inform perinatal education across Oceania, even as the original Tokyo meeting was reorganised because of pandemic-related travel restrictions. Within that broader perinatal conversation, neonatal anaemia remains one of the most frequent reasons a preterm baby receives a blood product during the first weeks of life.
The fetal liver is responsible for erythropoietin production rather than the kidney, and the resulting hormonal output is sluggish compared with adult physiology. A newborn delivered at 28 weeks has a small circulating volume, often under 100 mL per kilogram, and may lose 0.5 to 1 mL per day through routine blood sampling to monitor electrolytes, bilirubin and glucose. Over two or three weeks these losses add up to a significant proportion of the baby's total red cell mass, particularly when central lines are required for parenteral nutrition.
Compounding the picture, preterm infants have a shortened red cell lifespan of roughly 60 to 80 days instead of the 120 days seen in healthy adults, and they cannot adequately increase iron absorption to keep pace with the limited erythropoiesis that does occur. The result is the classic anaemia of prematurity, which typically peaks between four and eight weeks of life, but may appear earlier in the smallest babies. Clinicians recognise this window and plan feeding protocols, iron supplementation and transfusion thresholds accordingly.
Recombinant human erythropoietin has been used in neonatal units across Australia for several decades, with protocols varying between tertiary centres such as the Royal Women's Hospital in Melbourne, the Mater Mothers' Hospital in Brisbane and the Women's and Children's Hospital in Adelaide. The standard approach has shifted from subcutaneous epoetin alfa three times weekly to longer-acting agents such as darbepoetin alfa given once weekly or every ten days, which reduces handling of fragile infants and fits better with busy nursing rosters.
The rationale for erythropoietin therapy is to stimulate endogenous red cell production, reduce donor exposure and avoid the well-documented complications of transfusion. Contemporary Australian protocols generally start treatment in the first week of life for infants below 30 weeks' gestation, dosing between 200 and 400 units per kilogram of epoetin alfa or 10 µg per kilogram of darbepoetin, alongside enteral iron at 2 to 6 mg per kilogram per day once feeds are tolerated. The PBS schedule lists certain presentations, and many units participate in local audits to track transfusion rates before and after introducing routine early erythropoietin.
Important caveats include the historically observed association between early high-dose erythropoietin and retinopathy of prematurity, which prompted more conservative dosing in many Australian centres. There is also ongoing interest in using erythropoietin for neuroprotection at higher doses than those used for anaemia, although this remains a research application rather than standard care. For clinicians seeking background on related perinatal topics, the corticosteroid therapy resource on the same FAOPS platform offers useful context on another intervention used in the same population.
While erythropoietin aims to reduce the need for transfusion, red cell transfusion remains the fastest way to restore oxygen-carrying capacity in an unstable infant. The central question is no longer whether transfusion works, but when it is worth the immunological and metabolic risks. Australian NICUs generally follow a restrictive approach, applying transfusion thresholds that vary with respiratory support, gestational age and postnatal age.
A widely used framework in this country categorises infants by whether they are ventilated, on non-invasive respiratory support, or breathing spontaneously. A ventilated baby may receive transfusion at a haemoglobin threshold of around 110 to 120 g/L in the first week, falling to around 90 g/L later, while a baby on nasal continuous positive airway pressure may be transfused at 90 to 100 g/L. Spontaneously breathing preterm infants often tolerate thresholds of 70 to 80 g/L provided they are growing and have adequate reticulocyte responses. These numbers are derived from landmark trials such as the Iowa, PINT and ETTNO studies and have been incorporated into National Blood Authority guidance.
Restrictive and liberal transfusion strategies reflect different philosophies of when donor blood is justified, and the table below sets out the practical differences for a preterm infant.
| Feature | Restrictive strategy | Liberal strategy |
|---|---|---|
| Typical haemoglobin threshold | Lower, often 70–90 g/L in stable infants | Higher, often 100–120 g/L |
| Average number of transfusions per infant | Reduced by 1–3 over admission | Higher, with greater cumulative donor exposure |
| Primary goal | Minimise donor exposure, conserve blood supply | Optimise oxygen delivery at all times |
| Short-term physiological effect | Comparable tissue oxygenation in most stable infants | Modest improvement in cardiac output and oxygen delivery in unstable infants |
| Neurodevelopmental outcomes at 2 years | Non-inferior in ETTNO and larger meta-analyses | Similar to restrictive in most analyses |
| Suitability for ventilated infants | Higher thresholds applied when FiO₂ is elevated | Lower thresholds less common |
| Suitability for stable growing infants | Preferred approach in Australian practice | Rarely used outside specific protocols |
The table summarises the dominant Australian approach, which combines restrictive thresholds for stable infants with higher triggers for those on significant respiratory support. Local data from the Australian and New Zealand Neonatal Network suggest that adoption of restrictive thresholds has coincided with a measurable drop in transfusion rates without an increase in adverse neurodevelopmental outcomes reported at two years.
The risks of transfusion in neonates include transfusion-transmitted cytomegalovirus, although Australian donor screening and leucodepletion make this rare. Other recognised complications are volume overload, electrolyte shifts, hypocalcaemia from citrate and the risk of transfusion-associated gut injury in preterm infants with evolving necrotising enterocolitis. Donor exposure matters because each transfusion introduces a new unit, and most Australian centres now allocate single-donor paediatric packs to limit this exposure for the smallest babies.
Reducing phlebotomy losses is one of the most practical steps a unit can take. Bedside point-of-care analysers, which are widely used in tertiary NICUs in Sydney and Melbourne, can halve the volume of blood drawn each day when compared with sending samples to a central laboratory. Clustering blood tests, using closed microtainer systems and reserving routine daily sampling for unstable infants are simple measures that complement any erythropoietin protocol. When phlebotomy is minimised and iron supplementation is optimised, the infant's own marrow often produces enough red cells to delay or avoid transfusion altogether.
Care for extremely preterm infants is concentrated in a small number of tertiary perinatal centres, and regional or remote births are typically transferred via NETS, the Newborn and paediatric Emergency Transport Service. This means transfusion decisions are usually made by experienced neonatologists rather than general paediatricians, which supports consistent application of evidence-based thresholds. Parents are increasingly involved in these conversations, particularly through family-integrated care models now embedded in several Australian nurseries, and they often ask about donor exposure, breastfeeding and iron supplementation before discharge.
On the broader policy side, Australia relies on the Australian Red Cross Lifeblood service for donor blood, and the National Blood Authority oversees product safety, inventory and product development. The Therapeutic Goods Administration regulates biological therapies including recombinant erythropoietin products, while the Pharmaceutical Benefits Scheme determines subsidy for outpatient iron preparations used after neonatal discharge. Together these agencies create a tightly regulated environment in which both erythropoietin and transfusion are used judiciously, audited and continuously reviewed against the best available evidence from Australia and overseas.
Discuss the practical application of these tools with your neonatal team at the next ward round, review your unit's transfusion thresholds against current National Blood Authority guidance, and consider whether phlebotomy reduction initiatives could trim donor exposure for your smallest patients.