Ion-selective electrodes for sodium and potassium: a new problem of what is measured and what should be reported.

For clinical purposes the activities of Na+ and K+ obtained with ion-selective electrodes in undiluted whole blood or serum should be multiplied by an appropriate factor to obtain the same values as the substance concentrations obtained by flame photometry. The factor is primarily dependent on the mass concentration of water in normal plasma divided by the molal activity coefficient of Na+ (or K+) of normal plasma. We discuss the value of the molal activity coefficient of Na+ obtained by theoretical calculations and by direct measurement. The discrepancies between theory and measurement (gamma Na+ of 0.747 and 0.73, respectively) may be due to some binding of Na+ (protein binding or ion pair formation), a small and variable residual liquid-junction potential, or certainty about the appropriate value for the ionic strength of normal plasma (0.16 mol/kg or somewhat higher).