Abstract

The aim of the study was to compare the results of digoxin assays performed using fluorescence polarization immunoassay (FPIA) on the TDxFLx^® and a dry chemistry enzyme immunoassay (EIA) on the Vitros 950. Within-run CV amounted to 8.52–10.49% for FPIA and 2.47–5.39% for EIA. Between-run CV amounted to 6.41–8.97% for FPIA and 3.40–5.04% for EIA. Analytical bias ranged from 2.57–4.0% for FPIA and from 9.86–11.9% for EIA. In comparative studies the correlation coefficient was 0.878; Deming regression analysis yielded a slope of 1.057 (95% CI: 0.573 to 1.541) and intercept of 0.078 (95% CI: –0.391 to 0.547), and the Passing-Bablok agreement test yielded a slope of 1.111 (95% CI: 0.988 to 1.212) and intercept of 0.094 (95% CI: –0.018 to 0.182). The mean digoxin concentration in patients’ sera measured by EIA was significantly higher than that measured by FPIA (1.347 vs. 1.196 ng/ml, p < 0.02). The mean absolute difference between results amounted to 0.146 ng/ml (95% CI: 0.0261 to 0.266). In comparison to EIA, FPIA yielded a higher number of subtherapeutic concentrations < 0.5 ng/ml (29.7% vs. 21.8%) and a lower number of digoxin concentrations > 1.2 ng/ml (25.7% vs. 35.6%). These discrepancies occurred in approximately 10% of samples. The obtained results showed different analytical performance and method-dependent differences in the distribution of results. This indicates the necessity to harmonize digoxin immunoassays if two different analytical systems are used in the same clinical setting.