EJECTION FRACTION BY TRANSTHORACIC ELECTRICAL BIOIMPEDANCE REPRODUCIBILITY AND COMPARISON WITH MULTI-GATED RADIONUCLIDE SCAN

Reproducibility of ejection fraction (EF) by transthoracic electrical bioimpedance (TEB) was investigated in normal male volunteers. In a sh ort-term reproducibility study with 6 volunteers, the coefficient of v ariation was 2% within and between time points. Day to day reproducibi lity was studied in 25 volunteers on 2 separate days. TEB EF was repro ducible from day to day under various physiological stresses, with the coefficient of reproducibility varying from 5% (during supine rest) t o 13% (at 2 min post-exercise). In a separate study comparing TEB EF w ith multi-gated radionuclide imaging (MUGA) in 35 consecutive patients with cardiac disease, agreement [(MUGA-TEB) bias=-7 +/- 19%, p = 0.02 6] and correlation (r = 0.16) between the 2 methods was poor. Linear r egression analysis did not show significant correlation between TEB sy stolic time ratio (STR) and MUGA EF (F=3.13, p=0.093, r=0.38). We conc lude that although TEB EF is reproducible in the short-term and from d ay to day in healthy volunteers, it is a measurement of STR rather tha n the true EF. As the STR and true EF may alter in different ways unde r certain circumstances, TEB EF data can be misleading, and this param eter has to be interpreted with care in clinical pharmacology studies in such subjects. Furthermore, since TEB can provide STR data, there i s not advantage in transforming these data into EF. TEB cannot reliabl y predict the MUGA EF in clinical practice due to various factors conf ounding the relationship between STR and EF, and therefore cannot be u sed to routinely measure EF in patients.