Plasma samples (n = 155) of 30 patients on an intensive ward were analysed for magnesium simultaneously by atomic absorption spectrophotometry (AAS) and methylthymol blue spectrophotometry. Methylthymol blue spectrophotometry was performed at the bedside, using two different multianalysers, Easy ST 1 and Easy ST 2, Merck, D-Darmstadt. Precision was 12.2% (Easy ST 1) and 17.1% (Easy ST 2), and the average value was 0.89 mmol/l, which was above the expected range (0.72-0.88 mmol/l). Accuracy was 16.25% (Easy ST 1) and 8.75% (Easy ST 2). Analyser 2 was more accurate (8.75% versus 6.25%) but less precise (17.1% versus 12.2%) than analyser 1. Precision of AAS was between the expected values of 0.69 and 0.84 mmol/l. Easy ST and AAS gave significantly different values (p < 0.0001) for 155 measurements. Comparison of AAS and methylthymol blue spectrophotometery showed that methylthymol blue spectrophotometry produced higher values than AAS (mean difference 0.186 mmol/l). Furthermore, analyses of 40 samples of a standardized plasma concentration with methylthymol blue spectrophotometry showed a very low precision (15.3%). Easy ST cannot be assigned for urinary measurements of magnesium. Experimentally measured samples gave unaccountable results.
