Analysis of concordance between the bioelectrical impedance vector analysis and the bioelectrical impedance spectroscopy in haemodialysis patients

Analysis of concordance between the bioelectrical impedance vector analysis and the bioelectrical impedance spectroscopy in haemodialysis patients impedance spectroscopy (MF-BIS) are different and not comparable. Objective: Analyse whether the inter-method variability is due to bioelectrical variables measured by the different monitors, or rather due to the equations used to calculate body volume and mass. Another objective was to determine whether, despite the inter-method variability, the classification of hydration status by the two methods is consistent. Material and Methods: Bioelectrical impedance was measured by SF-BIVA and MF-BIS immediately before a dialysis session in 54 patients on haemodialysis. In 38 patients, the study was repeated by SF-BIVA at the end of the same dialysis session. Results: Resistance and phase angle values provided by the two monitors at a frequency of 50kHz were consistent. For resistance, variability was 1.3% and the intra-class correlation coefficient was 0.99. For phase angle, variability and the intra-class correlation coefficient were 11.5% and 0.92, respectively. The volume values for total body water, extracellular water, fat mass and body cell mass were biased, with a level of variability that would not be acceptable in clinical practice. The intra-class correlation coefficient also suggested a poor level of agreement. SF-BIVA systems define overhydration or dehydration as a vector below or above the tolerance ellipse of 75% on the longitudinal axis. MF-BIS uses two criteria for pre-dialysis hyper-hydration: overhydration (OH) greater than 2.5 litres, or greater than 15% of extracellular water. The degree of equivalence with the results of the SF-BIVA monitor was better with the second criterion (kappa: 0.81, excellent agreement) than with the first one (kappa: 0.71, acceptable agreement). The MF-BIS system defines post-dialysis normal hydration as a difference between OH and ultrafiltratation volume between -1.1 and 1.1 litres and agreement with the SF-BIVA system for this parameter was acceptable (weighted kappa index: 0.64). Conclusions: The MF-BIS and SF-BIVA systems provide similar readings for bioelectrical parameters, and the wide variation in the quantification of volume and body mass must be attributed to the different equations used for calculation. Furthermore, the criteria used by both systems to define both pre- and post-dialysis hydration have an acceptable level of equivalence.