ABSTRACT
OBJECTIVES: The objective of this study was to determine the in vivo correlation of ionized magnesium (iMg) with ionized calcium (iCa), total calcium, albumin and pH. In addition, the analytical interference of iCa on iMg measurement on the Stat Profile Prime Plus (Nova Biomedical) and vice versa was defined. METHODS: In vivo correlation of iCa, iMg and pH was studied in 238 paired blood gas samples of 109 different patients admitted to the intensive care unit. Albumin and total magnesium (tMg) were measured in heparinized plasma samples. Measurement of iMg was performed with the ion selective magnesium electrode (ISE) of the Stat Profile Prime Plus (Nova Biomedical) and iCa and pH were measured with a Rapid Point 500 blood gas analyzer (Siemens). Albumin, total calcium and total magnesium were analyzed with a Siemens Atellica CH. Analytical interference of iCa with iMg and vice versa was investigated using unbuffered saline solutions. RESULTS: In the studied patient population, no significant correlations were observed between iMg and iCa, albumin, and pH. An inverse relationship was observed between iCa and Mg-ISE. For every 0.1 mmol/L change in iCa concentration, the iMg concentration deviated by 0.01 mmol/L at an iMg concentration of 0.5 mmol/L and by 0.013 mmol/L at an iMg concentration of 1.0 mmol/L. The measurement of iCa was not affected by iMg. CONCLUSIONS: In vivo, no correlation was observed between iMg with iCa, albumin and pH. Interference of iCa on iMg measurement was noted, with a maximum deviation of ±0.02 mmol/L iMg across the reference range of iCa (1.15-1.32 mmol/L). Additionally, the iCa measurement was not affected by the iMg concentration.
ABSTRACT
Background: We evaluated the performance of the Abbott N-terminal pro-brain natriuretic peptide (NT-proBNP) assay against the Roche NT-proBNP immunoassay across two sites. Methods: Precision, linearity, and sensitivity studies were performed. A combined method of comparison and regression analysis was performed between the Roche and Abbott assays using samples from both sites (n = 494). To verify biotin interference, lyophilised biotin powder was reconstituted and spiked into serum samples at two medical decision levels (final concentration 500/4250 ng/mL) and compared to controls. NT-proBNP was also measured in anonymised leftover sera (n = 388) in a cardio-renal healthy population and stratified into three age bands<50 (n = 145), 50−75 (n = 183) and >75 (n = 60). Results: Between-run precision (CV%) for NT-proBNP was 4.17/4.50 (139.5/142.0 pg/mL), 3.83/2.17 (521.6/506.3), and 4.60/2.51 (5053/4973), respectively. The assay was linear from 0.7−41,501 pg/mL. The limit of blank/quantitation was 1.2/7.9 pg/mL. The assay showed no interference from biotin up to 4250 ng/mL. Passing−Bablok regression analysis showed excellent agreement between the two assays (r = 0.999, 95% CI 0.999 to 0.999, p < 0.0001). The Roche assay had a slightly persistent, negative bias across different levels of NT-proBNP. ESC age cut-offs for diagnosing acute heart failure are applicable for the Abbott assay, with the median NT-proBNP of subjects < 50 years old at 43.0 pg/mL (range 4.9−456 pg/mL), 50−75 years old at 95.1 pg/mL (range 10.5−1079 pg/mL), and >75 years old at 173.1 pg/mL (range 23.2−1948 pg/mL). Conclusions: The Abbott Architect NT-proBNP assay has good performance that agrees with the manufacturer's specifications. ESC/AHA recommended NT-proBNP age groups for acute heart failure diagnosis are applicable to this assay.