An assessment of the accuracy of renal blood flow estimation by Doppler ultrasound

Background: Knowledge of renal blood flow is considered important in the management of critically ill patients with acute renal failure. Renal Doppler ultrasound has been used to estimate renal blood flow. Its accuracy, however, has not been formally assessed. Design: Prospective blinded animal study. Setting: University physiology laboratory. Subjects: Seven merino cross-ewes. Interventions: We chronically implanted transit-time flow probes around the left renal artery and performed Doppler ultrasound measurements of RBF. We simultaneously recorded RBF values obtained with implanted flow probes and Doppler ultrasound during (a) observation, (b) dobutamine and (c) nitroprusside infusion in random order. Results: In a total of 202 paired measurements, Doppler ultrasound measured peak systolic velocity (PSV) correlated very weakly with implanted flow probe measurements of RBF (r ² = 0.015), as did end-diastolic velocity (EDV; r ² = 0.086) and mean velocity (MV_vel; r ² = 0.04). We also found similar weak correlations with other Doppler-ultrasound-derived indices. All comparisons showed bias and wide limits of agreement. Conclusions: Doppler-ultrasound-derived estimates of RBF show little correlation with transit-time flow probe measurements, display significant bias and wide limits of agreement and have low accuracy for clinically significant changes in RBF in large animals.