Clinical evaluation of sodium ion selective field effect transistors for whole blood assay

Sodium ion selective field effect transistors (ISFETs) were evaluated for their performance in measurement of sodium ions in whole blood for 'near patient' analysis in operating theatres and intensive care units. Performance was evaluated in comparison with a standard clinical laboratory sodium/potassium ion analyser (Radiometer KNA1) and with sodium and potassium assays using flame photometry on the plasma from each whole blood specimen. The imprecisions (coefficients of variation) of three ISFETs for sodium ion assay were 1.08, 1.56 and 1.10%, respectively. Robust bivariate linear regression (reweighted least squares preceded by least median of squares) of the ISFET versus KNA1 sodium ion activity yielded a regression coefficient of 1.08 and an intercept of -18.2 mM. The influence of potassium, protein and lipid on the measurement of sodium ions by both ISFETs and the KNA1 was assessed using robust multiple regression (also based on reweighted least squares preceded by least median of squares). In the regression versus flame photometry, protein was found to be more influential for the KNA1 (glass sodium ion selective electrode) than for the ISFET. Potassium had no influence on assays using the ISFET, but had a weak negative influence on assays using the KNA1. Two ISFETs lasted for more than 200 assays each demonstrating their robustness in the assay of whole blood.