Temperature-dependent matrix effect in the direct enzymatic measurement of blood glucose

The influence of the high-molecular-mass sample matrix in the direct enzymatic measurement of glucose in haemolysate was investigated by a comparison study using ultrafiltered haemolysate for reference. Haemolysate was obtained by 1:21 dilution of whole blood with a solution of digitonin and maleinimide. It was shown that at low protein concentration glucose distributes in a 1:1 ratio during ultrafiltration. With a hexokinase/glucose-6-phosphate dehydrogenase procedure excellent agreement was found between values measured in haemolysate (y) and ultrafiltrates (x), when incubation was performed at 25 degrees C (a = 0.047 mmol/l; b = 0.99; r = 0.999, n = 37); at 37 degrees C, however, the same procedure resulted in a non-tolerable systematic deviation in the direct analysis of haemolysate (a = -0.426 mmol/l; b = 1.00, r = 0.997, n = 37). The precision of measurements in haemolysate and ultrafiltrate was similar (CV 1.0-1.2%). Since stable reference material with an appropriate matrix is not available, it is important to evaluate haemolysate procedures carefully by comparison studies with patient samples. For reduction of experimental error in such studies we recommend the use of ultrafiltered haemolysate, since this can be analysed side by side with haemolysate in the same run.