Original Article

The Effects of Hematocrit, Amplifier Input Impedance and Magnet Excitation Frequency on the Accuracy of Electromagnetic Flowmeters

Ohio State University Columbus, OH Extracorporeal Technologies, Inc. Indianapolis, IN

^* Direct communications to: Philip A. Wagoner, 428 Porter Road, Charleston, WV 25314

Abstract

Since the introduction of electromagnetic flowmeters, clinical users have found it necessary to contend with several problems in the accuracy of the flow signal. The cumulative results of these inaccuracies has commonly been referred to as the “hematocrit effect”.

Blood was pumped through an in vitro circuit containing a segment of canine femoral artery. The hematocrit of the blood was varied from 10% to 70%. Flow through the artery segment was measured with two clinically popular electromagnetic flowmeters employing perivascular probes. Measured flow was compared to actual flow to quantitate the “hematocrit effect”.

Both flowmeters increasingly overindicated flow as the hematocrit was decreased (r = .85, p < .01, m =- 1.51). The overindication of flow was a result of entering a manufacturer’s precalibration constant that inaccurately corrects for voltage assumed lost to erythrocyte impedance.

It is recommended that electromagnetic flowmeters not be used to evaluate coronary artery graft flow unless adequate compensation for hemodilution is made.

The amplifier input impedance and magnet excitation frequency were found to be great enough to prevent detraction from the flowmeter’s accuracy at all hematocrits studied (r > .90, p < .01).

In an ancillary study, the Carolina Medical Electronics, Inc. model 701-D Cliniflow IF^™ flowmeter was found to be independent of the “hematocrit effect” (all rs > .98, all ps < .01). The Cliniflow II^™ has a “user dictated” microprocessor that accurately compensates for the hematocrit effect.