Proposal for standardizing direct-reading biosensors for blood glucose

Abstract

Direct-reading glucose biosensors sense molality (glucose per unit water mass) in the sample. With aqueous calibration, a direct-reading glucose biosensor produces higher results in blood and plasma than methods measuring concentration, theoretically by the ratio of water concentrations in calibrator and sample. To confirm this, we measured glucose in 14- blood and 40 plasma samples with the direct-reading glucose sensor in the Chiron Model 860 Blood Gas and Critical Analyte System and with our routine method (ESAT 6660; Eppendorf). The Chiron instrument is calibrated with a 10 mmol/L (180 mg/dL0 glucose calibrator (mass concentration of water = 0.99 kg/L). Assuming normal water concentrations of 0.84 and 0.93 kg/L in blood and plasma, respectively, we multiplied results from the Chiron sensor by 0.84/0.99 and 0.93/0.99 to obtain concentrations in blood and plasma. This conversion resulted in agreement of results with our routine method. An individual conversion based on hematocrit in each whole-blood sample was less satisfactory. To avoid confusion over variously measured and reported glucose results and reference values, we suggest standardization and reporting of whole-blood glucose results as equivalent plasma concentrations. This proposal may be conveniently achieved by using a commercially available reference material for glucose, NIST SRM 965.