[Measurement of extravascular lung water: a toy or tool?].

We examined the reproducibility of the results of extravascular lung water measurements obtained by the double indicator dilution method. The coefficient of variation (delta) for 82 EVLW-measurements was 13 +/- 8%. The delta of the 3 variables measured and used to calculate the EVLW value i.e. cardiac output, mean transit time for dye (MTTD) and mean transit time for the thermal signal (MTTT) were all below 10%. Our results suggest that the important coefficient of variation of the EVLW-measurements is overall due to the variability of the difference between the 2 transit times measured. The temperature exchange between the intravascular cold bolus and the extravascular thermal volume is flow dependent especially at high EVLW-values.