Validation of the measurement of haemodialysis access flow using a haemoglobin dilution test

Monitoring of blood flows in arteriovenous fistulae and arteriovenous grafts is recommended to predict access thrombosis. The ultrasound dilution technique (UDT) is the gold standard. We compare a recently described haemoglobin dilution technique (HDT) with the UDT in measurement of vascular access flow. Access blood flow was measured in 67 stable dialysis patients using HDT by bedside Hemocue (Hemocue AB, Ängelholm, Sweden) and laboratory measurement. Access blood flow was then measured by UDT in the same dialysis session. Median flow rate by UDT was 950 ml/min (IQR 490-1,440 ml/min), by Hemocue HDT 935 ml/min (IQR 475-1,395 ml/min, p = 0.534), and by laboratory haemoglobin HDT 920 ml/min (IQR 463-1,378 ml/min). Bland-Altman plots demonstrated poor agreement between UDT and HDT (limits of agreement for Hemocue HDT -22.7 to 20.1%, for laboratory HDT -21.2 to 20.4%). HDT can be used to measure vascular access flow but requires validation against clinical outcomes before being recommended as an alternative to UDT.