RESUMO
BACKGROUND: The Glucose Pump Test (GPT) is a new method for measuring the access blood flow (Qa) based on a constant glucose infusion (Qi), with known glucose concentration (Ci), in the arterial needle and on glucose determination in two blood samples from the venous needle, the first (C1) in basal conditions, the second (C2) during the infusion. Qa depends on the difference in glucose between the two samples and is computed from the formula: QaGPT = Qi x (Ci-C2)/(C2-C1). METHODS: The new method, previously evaluated by ultrasound dilution and color-Doppler techniques, was compared with the Urea Test (UT) in 20 patients measuring recirculation (R) during reversal of the arterial and venous needles (QaUT = Qb x (1/R -1)). All Qa determinations were done twice by both methods. Glucose and urea were determined respectively two and three times. RESULTS: Mean QaGPT = 841, SD 347 ml/min, mean QaUT = 872, SD 417 ml/min (p = n.s.); mean percent difference QaGPT-QaUT= 16%, SD 14; mean coefficients of variation of paired determinations: 8.1% and 12.1% respectively; Pearson coefficient between the two methods: r= 0.91. CONCLUSIONS: The comparison showed a good correlation between the two methods and similar mean values. The coefficient of variation of the new method was acceptable and lower than with the UT. The GPT is a reliable technique for measuring blood flow in vascular accesses.