Evaluation of five in-line hematocrit monitors

Monitoring the hematocrit is essential during cardiopulmonary bypass for efficacious administration of blood products. The purpose of this study was to evaluate five cardiopulmonary bypass in-line monitors designed to display continuous hematocrit or hemoglobin values. The devices were evaluated for accuracy using an in vitro circuit primed with human blood while randomizing hematocrit, blood flow rate, and temperature. Hematocrits correlated significantly with the error in all the devices (p < 0.01). Over evaluation time, the error of the CDI, Gish, IBC, and MX2 increased significantly (p < 0.05). Temperature correlated significantly with the error of the Gish device (r = -0.49, p < 0.01). Blood flow correlated significantly with the Gish error (r = -0.24, p < 0.01). The Cobe device had a significantly smaller overall error than the other devices (p < 0.001). Device evaluation, based on a low mean error, a low percent error, a high correlation with the actual hematocrit, low correlations between mean error, blood flow, and temperature, and insignificant correlation between time and error, suggests that the Cobe device is more accurate for the continuous monitoring of hematocrit during cardiopulmonary bypass.