RESUMO
We evaluated four commercial radioimmunoassay kits for digoxin. We assayed a standard plasma containing digoxin, 2.0 microgram/L, and samples from patients receiving digoxin, with use of the kits and of a bioassay, the 86Rb-uptake inhibition technique. Intra-assay precisions differed significantly. Computer-calculated 95% confidence intervals for the radioimmunoassays averaged 0.4 to 0.6 microgram/L at the proposed toxic threshold of 2.0 microgram/L; the corresponding value of the 86Rb assay was 0.75 microgram/L. Digoxin in the standard plasma was overestimated with three of the kits (means: 2.40, 2.56, and 2.59 microgram/L) but was assayed accurately by the 86Rb technique and by one kit. This same kit gave a significantly lower mean (1.07 microgram/L) for the patients' samples then did the other three kits (1.32, 1.49, and 1.29 microgram/L), two of which also differed significantly in accuracy. The 86Rb assay measured glycoside activity corresponding to a mean digoxin concentration of 1.35 microgram/L. We conclude that the relatively low precision of digoxin assay and the variations in accuracy between kits from various vendors apparently deserve continual attention.