Hemodialysis access flow measurement: Comparison of ultrasound dilution and ultrafiltration method on NIKKISO DBB-EXA™ dialysis machine

Background: The methods of estimating vascular access (VA) flow rates are usually based on the indicator dilution theory by measuring recirculation during dialysis sessions. Methods: This is an observational study comparing the VA flow rates measured by NIKKISO DBB-EXA (TM) and Transonic (R). Sixty-five patients (38 M/27 F, mean age 72 +/- 10 years) participated in the study. We measured the VA flow rates during dialysis twice with each method and repeated the procedure 7 days later. Results: In 130 double measurements for each method on the same day, mean flow with Transonic (R) was 1413 +/- 715 ml/min and with DBB-EXA (TM) 1297 +/- 664 ml/min. In Bland-Altman analysis, the mean difference between the two methods was 159 +/- 211 ml/min (limits of agreement: -274 and 572 ml/min). Eighty-one out of the 130 DBB-EXA (TM) measurements were within 25% of the Transonic (R) measurements (62% accuracy). Regarding reproducibility of each method on different days, mean difference in the Bland-Altman analysis was 29 +/- 620 ml/min (limits of agreement: -1186 and 1244 ml/min) for the Transonic (R) measurements and 132 +/- 625 ml/min (limits of agreement: -1092 and 1356 ml/min) for the DBB-EXA (TM) measurements. The measurements on two different days were within 25% of each other for 52 of the 65 patients (80%) with the Transonic (R) method, and for 35 of the 65 patients (54%) with the DBB-EXA (TM) method. Conclusions: In conclusion, the DBB-EXA (TM) method underestimates VA flow rates compared to the Transonic (R) technique, resulting in a limited accuracy of 62%. There was poor reproducibility for both methods in different day measurements with better performance of the Transonic (R) technique. The DBB-EXA (TM) method could be used as a simple tool for a rough estimate of VA flow rates but cannot replace the Transonic (R) reference method.