Background: Whether the thoracic fluid content (TFC) estimated by bioimpedance actually reflects the thoracic water content and tracks its changes has not been tested yet. We compared TFC changes induced by volume expansion with those of well-established markers of cardiac preload including the global end-diastolic volume indexed (GEDVI) estimated by transpulmonary thermodilution (TPTD) and central venous pressure (CVP). We assessed the relationship between TFC and extravascular lung water indexed (EVLWI) estimated by TPTD in patients with acute respiratory distress syndrome (ARDS).
Methods: In 42 patients with a TPTD device (PiCCO2), we measured GEDVI, EVLWI and CVP before and after a 500-mL fluid bolus (Fluid group). In 23 patients with ARDS, we measured the day-to-day changes in TFC and EVLWI (ARDS group).
Results: In the Fluid group (42 measurements), TFC increased significantly during fluid infusion. There was no correlation between the fluid-induced changes in GEDVI or CVP and the fluid-induced changes in TFC (p=0.71 and p=0.09, respectively). In the ARDS group, 124 measurements were performed, representing 101 changes between timepoints. There was no correlation between the percent changes in GEDVI, EVLWI, the sum of GEDVI+EVLWI and the percent changes in TFC (p=0.52, 0.40 and 0.55, respectively). Considering all 208 pairs of measurements performed, there was no correlation between the absolute values of GEDVI, EVLWI and the sum of GEDVI+EVLWI and those of TFC (p=0.33, 0.04 and 0.39, respectively). Considering all pairs of measurements performed, the percentage of error of cardiac index measured by bioreactance compared to TPTD was 131%.
Conclusions: In critically ill patients, TFC measured by bioimpedance does not follow the changes induced by a fluid bolus of well-known markers of cardiac preload. It also does not follow the changes in EVLWI observed in patients with ARDS. It is determined neither by GEDVI, nor by EVLWI or the sum of the two.