Glucose pump test: a new method for blood flow measurements

Background. A good test for monitoring blood flow (Q(a)) must be accurate, rapid and economical in order to allow frequent easy measurements. The glucose pump test (GPT) is based on a constant glucose infusion as a dilutional indicator of Q(a). Methods. GPT protocol requires a constant glucose infusion, by a syringe pump, into the arterial needle and two blood withdrawals from the venous needle, one basal before the infusion (C-a1), the other (C-a2) 11 s after the start of the infusion. At the bedside we measure glucose on C-a1 and C-a2. Knowing the infused glucose concentration (C) and the pump infusion rate (Q(i)) we can easily calculate Q(a) = Q(i) x (C-i-C-a2)/ (C-a2-C-a1). We verified the accuracy of this new method by comparing it with the in vitro results from a circuit reproducing vascular access circulation, and in vivo comparing GPT-Q(a) with Doppler ultrasound in pre-dialysis to the Transonic HD01-Q(a) during dialysis in 23 chronic haemodialysis patients. Results. GPT-Qa values were highly correlated with the in vitro Q(a)=1.01xGPT-Q(a)-16.6; r = 0.94. There was agreement between the mean flow values of GPT and Doppler (927.5 and 927.1 ml/min, respectively; P = NS) while the mean value of HD01 was significantly lower (HD01-Q(a) = 690 ml/min; P < 0.001 vs GPT-Q(a) and Doppler-Q(a)). The regression analysis showed a good correlation between GPT and Transonic results (r = 0.95; HD01-Q(a) = 0.86 x GPT-Q(a)-111.9), while there was a significant difference between the two measurements (mean Delta 235 +/- 117 ml/min; range from 15 to 451 ml/min). This difference could be caused by the large haemodynamic variations (different blood pressure, cardiac output, circulating effective volume, haematocrit) between pre-dialysis and intra-dialysis and in addition by the counter current flow during the reversal blood lines Transonic measurements. Conclusions. GPT offers the advantage of a simple bedside procedure easily performed before dialysis: it does not interfere with the dialysis treatment and it is less intrusive for the patient as it does not involve reversal of the blood lines. The preliminary data indicate that our method could be a useful, simple and cheap test for monitoring access flow in every dialysis unit.