Neonatal Partial Exchange Transfusion for Polycythaemia Calculator
Neonatal Partial Exchange Transfusion for Polycythaemia Calculator
The volume to exchange with normal saline in a symptomatic polycythaemic newborn, from weight, the observed haematocrit and the target. The procedure is for symptomatic babies: the evidence in asymptomatic polycythaemia does not support it.
Partial exchange volume for polycythaemia
Weight + haematocrits → mL to exchangeA 3.2 kg baby, blood volume 90 mL/kg, haematocrit 70%, target 50%
Formula
where blood volume (mL) = weight (kg) × blood volume per kg
- blood volume per kg
- 80 to 85 mL/kg in the University of Iowa NICU handbook, 90 mL/kg in the UCSF Benioff consensus guideline, and about 100 mL/kg in a preterm baby. The sources differ, which is why this page asks rather than choosing — and why the answer should be checked against local protocol
- (observed − desired) ÷ observed
- the fraction of the circulating red cell mass that has to go. It falls out of the assumption that red cells are conserved: replacing a volume V of blood at haematocrit Hct_obs with the same volume of saline leaves the red cell mass reduced in exactly that proportion
- desired haematocrit
- 50% in the UCSF consensus guideline, 50 to 55% in the Iowa handbook. It must be below the observed value for the formula to mean anything
- 0.9% sodium chloride
- the replacement fluid. Isotonic crystalloid, not plasma or albumin — a randomised comparison found saline as effective, and it avoids the risks of a plasma-derived product
- what the formula assumes
- that the exchange is isovolaemic, that the haematocrit is uniform, and that no red cells are produced or destroyed during the procedure. None is exactly true, which is why the haematocrit is rechecked afterwards rather than assumed
- what it cannot tell you
- whether to do the exchange. That turns on whether the baby has signs attributable to hyperviscosity, and the evidence in the asymptomatic baby does not support the procedure
Worked example
A 3.2 kg baby, blood volume 90 mL/kg, haematocrit 70%, target 50%
Blood volume = 3.2 × 90 = 288 mL
Fraction to exchange = (70 − 50) ÷ 70 = 20 ÷ 70 = 0.2857
Volume = 288 × 0.2857 = 82 mL
That is about 29% of the baby's circulating volume, removed and replaced in aliquots of a few millilitres per kilogram with 0.9% sodium chloride
At 80 mL/kg instead of 90, the same baby needs 3.2 × 80 × 0.2857 = 73 mL — an 11% difference that comes entirely from which published blood volume figure is used, which is why this page names the source of each
A 3.5 kg baby with a haematocrit of 75% and the same 55% target needs 3.5 × 90 × (20 ÷ 75) = 84 mL
If the haematocrit were 58% and the target 50%, the volume would be 288 × (8 ÷ 58) = 40 mL — and at that distance from target the question is whether an exchange is warranted at all rather than how much to exchange
Exchange volume by haematocrit, engine arithmetic
| Weight | Blood volume per kg | Observed Hct | Desired Hct | Volume (mL) |
|---|---|---|---|---|
| 3.2 kg | 90 mL/kg | 70% | 50% | 82 |
| 3.2 kg | 80 mL/kg | 70% | 50% | 73 |
| 3.2 kg | 90 mL/kg | 70% | 55% | 62 |
| 3.2 kg | 90 mL/kg | 65% | 50% | 66 |
| 3.5 kg | 90 mL/kg | 75% | 55% | 84 |
| 2.0 kg | 90 mL/kg | 72% | 50% | 55 |
| 1.2 kg | 100 mL/kg | 70% | 50% | 34 |
What the evidence actually supports
| Situation | What the published guidance says |
|---|---|
| Symptomatic polycythaemia with signs attributable to hyperviscosity | Partial exchange transfusion is considered. Both source guidelines frame the indication around the presence of signs |
| Asymptomatic polycythaemia, or minor symptoms only | Cochrane: no proven clinically significant short or long-term benefit. Randomised trials in asymptomatic infants found no difference in neurodevelopmental outcome |
| Haematocrit above 65% on a capillary sample | Confirm on a peripheral venous or arterial spun haematocrit before acting. A capillary sample reads several points high |
| Haematocrit above 65% with signs | The point at which both guidelines say to consider the procedure — the number and the signs together, never the number alone |
| Risk of the procedure itself | Cochrane found an increased risk of necrotising enterocolitis in infants receiving partial exchange, relative risk 11.18 (95% CI 1.49 to 83.64) |
| Choice of replacement fluid | Isotonic crystalloid. A randomised comparison of saline against plasma found saline as effective, without the risks of a plasma-derived product |
What accompanies neonatal polycythaemia
| Finding | Why it matters |
|---|---|
| Hypoglycaemia | Common, and partly explained by glucose consumption by the increased red cell mass. Check the glucose before, during and after |
| Hyperbilirubinaemia | A larger red cell mass means more bilirubin as those cells are broken down. Jaundice frequently follows polycythaemia |
| Thrombocytopenia | Described in polycythaemic newborns, and the exchange itself dilutes platelets and clotting factors further |
| Respiratory distress, cyanosis, plethora, lethargy, jitteriness, poor feeding | The signs attributable to hyperviscosity — and the reason to consider an exchange at all |
| Growth restriction, post-term birth, maternal diabetes, delayed cord clamping, twin-to-twin transfusion, high altitude | The settings in which polycythaemia arises, and worth identifying because some of them change what else to look for |
| Capillary versus venous sampling | A capillary haematocrit from a warmed heel reads several percentage points above a simultaneous venous sample. This is a common reason for an apparently alarming result |
A volume that is easy to calculate and hard to justify
The arithmetic of a partial exchange transfusion is a conservation argument. If red cells are neither made nor destroyed during the procedure, and a volume of blood at the observed haematocrit is removed and replaced with an equal volume of saline, then the red cell mass falls in exactly the proportion of blood removed. Setting the final haematocrit to the target and solving gives the volume: the baby’s circulating blood volume multiplied by the observed haematocrit minus the desired one, divided by the observed. Every published version of the formula is that expression, and the only genuine disagreement between sources is the blood volume per kilogram — 80 to 85 millilitres in the Iowa NICU handbook, 90 in the UCSF consensus guideline, around 100 in a preterm baby. A ten per cent difference in that figure is a ten per cent difference in the volume, which is why it is offered here as a choice with the source named rather than silently fixed.
The harder question is whether the procedure should happen at all, and here the honest answer is uncomfortable for a calculator. Partial exchange transfusion is indicated for polycythaemia that is causing symptoms — respiratory distress, cyanosis, lethargy, jitteriness, poor feeding, plethora attributable to hyperviscosity. For the polycythaemic baby who is clinically well, the Cochrane review found no proven clinically significant short or long-term benefit, and randomised trials that enrolled only asymptomatic infants found no difference in neurodevelopmental outcome with or without the procedure. Worse, the review identified an increased risk of necrotising enterocolitis among infants who received an exchange, with a relative risk above eleven — a wide confidence interval on a small number of events, but pointing firmly in one direction. A treatment with unproven benefit and a signal of serious harm is not one to reach for because a number crossed a line.
So the threshold deserves careful wording. Both source guidelines say to consider partial exchange in an infant with a haematocrit above sixty-five per cent who is demonstrating signs. The haematocrit and the signs are joined, and the guidelines do not offer the haematocrit alone as a trigger. There is also a sampling trap sitting directly underneath it: a capillary haematocrit taken from a warmed heel reads several percentage points above a simultaneous venous sample, because of stasis in the peripheral circulation. A screening capillary result above sixty-five per cent is a reason to take a peripheral venous or arterial spun haematocrit, not a reason to prepare a procedure.
When an exchange is performed, a few practicalities matter. The replacement fluid is isotonic crystalloid — 0.9% sodium chloride — rather than plasma or albumin; a randomised comparison found saline as effective and it avoids the risks of a plasma-derived product. The exchange is done in aliquots with continuous cardiorespiratory monitoring, and the haematocrit is remeasured afterwards rather than assumed to have landed on target, because the assumptions of uniform haematocrit and perfectly isovolaemic exchange are approximations. And the things that travel with polycythaemia need following independently: hypoglycaemia is common and partly explained by glucose consumption by the expanded red cell mass, hyperbilirubinaemia follows the breakdown of that mass, thrombocytopenia is described and is made worse by dilution during the exchange, and abdominal signs and feed intolerance deserve attention for days afterwards given what the Cochrane review found.
Frequently asked questions
How is the partial exchange transfusion volume calculated in neonatal polycythaemia?
Volume in mL = blood volume × (observed haematocrit − desired haematocrit) ÷ observed haematocrit, where blood volume is the weight in kg multiplied by the blood volume per kg. For a 3.2 kg baby at 90 mL/kg with a haematocrit of 70% and a target of 50%: 288 × (20 ÷ 70) = 82 mL, exchanged with 0.9% sodium chloride.
What blood volume per kg should be used in a neonate?
The sources differ. The UCSF Benioff consensus guideline for partial exchange uses 90 mL/kg; the University of Iowa NICU handbook gives 80 to 85 mL/kg; a preterm baby is usually taken as about 100 mL/kg. Blood volume per kilogram is highest at birth and falls through childhood, so an adult-style figure understates it. Use your unit’s protocol figure.
Should asymptomatic neonatal polycythaemia be treated with partial exchange?
The evidence does not support it. A Cochrane review concluded there are no proven clinically significant short or long-term benefits in polycythaemic newborns who are clinically well or have only minor symptoms of hyperviscosity, and found an increased risk of necrotising enterocolitis in infants who received the procedure. The indication is framed around symptoms, not a haematocrit alone.
What target haematocrit is used?
The UCSF consensus guideline states the desired haematocrit should be 50%; the University of Iowa handbook gives 50 to 55%. The target must be below the observed value, or the formula returns a negative volume.
What fluid is used to replace the blood removed?
Isotonic crystalloid — 0.9% sodium chloride. A randomised comparison of saline against plasma found saline as effective, and crystalloid avoids the additional risks of a plasma-derived product. Blood is removed and an equal volume of saline infused, in aliquots rather than in one pass.
Why does a capillary haematocrit matter here?
Because it reads high. A haematocrit taken from a warmed heel is several percentage points above a simultaneous venous sample owing to peripheral stasis, so a screening capillary result above 65% should be confirmed on a peripheral venous or arterial spun haematocrit before any procedure is considered.
Related calculators
References
- UCSF Benioff Children’s Hospitals. Consensus Guidelines for Partial Exchange Transfusion for Polycythemia in Neonates — volume formula, 90 mL/kg estimated blood volume, desired haematocrit 50%, 0.9% sodium chloride as the exchange fluid.
- Dempsey EM, Barrington K. Short and long term outcomes following partial exchange transfusion in the polycythaemic newborn: a systematic review. Arch Dis Child Fetal Neonatal Ed. 2006;91(1):F2–F6.
- Dempsey EM, Barrington K. Crystalloid or colloid for partial exchange transfusion in neonatal polycythemia: a systematic review and meta-analysis. Acta Paediatr. 2005;94(11):1650–1655.
- Ozek E, Soll R, Schimmel MS. Partial exchange transfusion to prevent neurodevelopmental disability in infants with polycythemia. Cochrane Database Syst Rev. 2010;(1):CD005089 — no proven clinically significant benefit in well or minimally symptomatic infants; increased risk of necrotising enterocolitis.
- University of Iowa Stead Family Children’s Hospital. Hematology: NICU Handbook — estimated blood volume 80–85 mL/kg, desired haematocrit 50–55%.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.