Pre-clinical validation of a new intra-operative "dual beam Doppler" blood flowmeter in an artificial circuit.

BACKGROUND Intra-operative flow measurement during coronary or peripheral bypass operations is helpful for ruling out technical failures and for prediction of complication and patency rates. Preclinical validation of the flowmeters is required in order to rely on the intra-operatively measured results. The aim of this study is to evaluate a new "dual beam Doppler" blood flowmeter before clinical application and to compare it with the established "transit time flow measure-ment" technique in an artificial circuit. METHODS Measurements were performed in an experimental flow model using pig blood and pig arteries. Three different flowmeters were used: Quantix OR (dual beam doppler flowmeter), CardioMed (transit time flowmeter), and Transonic (transit time flowmeter). Three validation tests were performed to assess correlation, precision, and repeatability of devices. (1) Correlation and agreement analysis was performed with various flow amounts (10-350 mL/min) (n = 160). (2) Device reproducibility and measurement stability were tested with a constant flow (flow amount = 300 mL/min) (n = 30). (3) A user accuracy test (intra- and inter-observer variability) was performed by 5 different observers with a constant flow (flow amount = 205 mL/min) (n = 75). Time collected true flow was used as a reference method in all steps and all tests were performed in a blind manner. Results are shown as mean values +/- standard deviations. Pear-son's correlation and Bland-Altman plot analyses were used to compare measurements. RESULTS The mean flow was 167 +/- 98 mL/min for true flow and 162 +/- 94 mL/min, 165 +/- 94 mL/min, and 166 +/- 100 mL/min for Quantix OR, CardioMed, and Transonic, respectively. Correlation coefficients between Quantix OR, Medi-Stim, Transonic, and time collected true flow were over 0.98 (P = .01). Most of the measured results ( > 90%) were between +/- 1.96 SD agreement limits in Bland and Altman plot analysis. All devices showed good results in the reproducibility test. During the user accuracy test, larger variance changes were observed between intra- and inter-observer results with the dual beam Doppler flowmeter compared to the 2 used transit time flowmeters when used for single sided vessel access without stabilization device (available from the manufacturer). CONCLUSION All 3 tested flowmeters showed an excellent correlation to the true flow in an artificial circuit and the accuracy of the tested devices was within agreement limits. Reproducibility of all devices was good and linear. The new dual beam Doppler flow measurement technique compares favorably to the classic transit time method. Clinical use may depend on operator, location, and condition, thus more studies may be required to ensure uniform results using the currently available blood flow measurement devices.