Detection of ferricyanide as a probe for the effect of hematocrit in whole blood biosensors.

Measurement of the concentration of an analyte in whole blood can be influenced by a range of factors; the red cell content or hematocrit (Hct) of the sample, the distribution and rate of movement of analyte between red cells and plasma, the amount of protein in solution, the viscosity of the sample and fouling of the sensor. The effect of the red cells is the major factor that must be taken into account. Using the analyte molality rather than the analyte molarity, the theoretical response for a range of analytes which are found in plasma and in the red cells can be calculated. For an analyte which is found in plasma alone, the effect of hematocrit is significant, with a bias of -1% per %Hct; if the analyte can freely and rapidly diffuse between the red cells and plasma, this bias is reduced to zero. Using ferrocyanide as a model analyte, the effects of fouling and reduced sample viscosity were measured to be -0.2% per %Hct, giving an overall bias of -1.2% per %Hct, a level of bias which is not clinically acceptable. This bias can be negated by measuring the hematocrit separately and incorporating it into the measurement algorithm. Such a correction is essential for the correct measurement of the concentration of an analyte in whole blood.