A next generation enzymatic magnesium assay on the Abbott ARCHITECT chemistry system meets performance goals based on biological variation

Abstract

ObjectiveTo evaluate the performance of the Abbott ARCHITECT enzymatic assay for magnesium (3P68) in serum/plasma and urine against analytical goals based on biological variation. MethodsAnalytical performance was evaluated according to CLSI protocols. Precision was examined using commercial chemistry controls. Accuracy was assessed against NIST SRM 956c, electrolytes in human serum. Correlation with the arsenazo Mg assay (7D70) was completed using patient samples (plasma, N=101; urine, N=90). Common interferences were examined in pooled patient specimens with high and low magnesium concentrations. ResultsThe enzymatic Mg assay displayed imprecision of 1.7% at 0.72mmol/L and 1.4% at 1.80mmol/L (20days, one calibration, one reagent lot). The linear range was verified between 0.18–7.0mmol/L (plasma) and 0.01–10.69mmol/L (urine). Results of the enzymatic assay (x) correlated well with the predicate assay (y) with the relationships y=0.891x+0.035, R=0.967 (plasma) and y=1.181x+0.086, R=0.997 (urine). Mean bias of the NIST SRM 956c samples was −1.4%. This method showed minimal interference by hemoglobin (3g/L as hemolysate), lipemia (20g/L Intralipid), unconjugated bilirubin (531μmol/L), and ascorbate (680μmol/L). ConclusionsThe ARCHITECT Magnesium assay 3P68 achieved the desirable analytical quality specification of 4.8% for total allowable error. In comparison to the 7D70 assay, notable improvements are seen in precision, 30-day calibration stability, and minimal interference by hemolyzed and lipemic samples.