Abstract

Overall laboratory reproducibility for the Ames Seralyzer and plasma potassium test strips was less than 4% within the plasma reference range; and linearity extended from 2.1 to 10.2 mmol/l. Neither bilirubin nor lipaemia interfered in the analysis and selectivity for potassium over sodium was 450:1. Laboratory comparison with automated flame photometry and an indirect ion selective electrode showed a Seralyzer bias of +0.10 (+/- 25D 0.26) and +0.11 (0.27) mmol/l. Against flame photometry, comparison of analyses by six doctors from a coronary care unit and four nurses from a renal dialysis unit showed smaller positive biases (0.06 and 0.02 mmol/l) but substantially larger error ranges (0.43 and 0.55 mmol/l, respectively). Some 49% of the nurses' and 37% of the doctors' results differed from laboratory values by more than 0.2 mmol/l. Potassium test strip results proved acceptably comparable and reproducible when produced by an experienced analyst, but greater variability in clinical hands rendered the system unsuitable for precise monitoring of potassium concentration.