A Practical Analysis of the Electrical Conductivity of Blood

Recent developments in indicator-dilution measurement of pulmonary edema have generated new interest in the use of electrical-conductivity sensing for measurement of indicator concentrations in blood. This approach has always suffered from the lack of an appropriate and validated model of the dependence of electrical conductivity on blood composition and indicator concentration. Such a model is developed here, based in part on a review of earlier work on variation with hematocrit and on recognition of the profound effect on conductivity of even transient alterations of blood temperature by the indicator itself. Shifts of water into and out of erythrocytes, in response to osmotic pressures, are properly included. The model predicts approximately linear changes in blood conductivity'over a large range of indicator concentrations (e. g., independent nonlinearity of 2.6 percent for 3 percent saline over a 0-30 percent concentration in blood). Changes in resistivity can be nonlinear, particularly with hypertonic indicators (e. g., independent nonlinearity of 7.9 percent for 3 percent saline over a 0-15 percent concentration in blood). Overlooking the thermal and osmotic effects can lead to significant errors in signal interpretation; consequent errors in flow rate determinations can be greater than 10 percent. Model predictions are verified in vitro with various indicators and canine blood using a miniature tetrapolar conductivity cell of our own design. The multicomponent nature of most conductivity indicators and the associated implications are discussed.