Figure 1.
Schematic drawing of the lung model.
The lung and ventilation compartments are partly filled with distilled water, their volumes are 6.94 L and 11.72 L, respectively. The outer chamber's volume is 67.73 L, filled with distilled water and constantly heated to 37°C. The dashed line reflects the water level determining tidal volume and functional residual capacity (FRC) at in- and expiratory end-tidal levels. FRC is the volume of air contained in the model at end-tidal expiration. One mm on the tape measure corresponds to 18.7 mL volume. FRC range is 0.5–4.2 L. O2 = sampling point of the side-stream oxygen sensor. CO2 = sampling point of the main-stream carbon dioxide sensor. Flowhead = main-stream ultrasonic flowmeter.
Table 1.
Study protocol for the lung model.
Figure 2.
In vitro variability of functional residual capacity measurement.
Bland Altman plot [16] of measured functional residual capacity (FRC) minus lung model FRC plotted vs. mean of measured and model FRC. Differences (open circles), and mean difference, and upper and lower limits of agreement (mean difference ±1.96 SD of differences) are given as dashed lines for the small lung model and as solid lines for the large lung model. Figure 2a gives absolute FRC differences (mL). For the small and large lung model, mean differences are −10.2 mL and 12.5 mL, upper limits of agreement are 9.9 mL and 121.9 mL, and lower limits of agreement are −30.3 mL and −96.8 mL, respectively. Figure 2b gives relative FRC differences (%). For the small and large lung model, mean differences are −3.8% and 0.4%, upper limits of agreement are 3.6% and 4.7%, and lower limits of agreements are −11.2% and −4.0%, respectively.
Figure 3.
Association of measurement error with lung volume.
Functional residual capacity (FRC) measurement error (%) is plotted against lung model FRC. The error data are displayed as open circles for the small model (100–400 mL FRCs) and closed circles for the large model (500–4200 mL FRCs). The dashed line gives the 5% limit of acceptable measurement error, the dotted lines reflect possible nominal volume bias due to reading parallax error (one mm) in the small and large lung model.
Table 2.
Linearity of sensors.
Table 3.
Nitrogen multiple-breath washout outcomes and variability.
Figure 4.
In vivo variability of functional residual capacity.
Bland Altman plot [16] of functional residual capacity (FRC) measured on two study days within three weeks in 30 healthy adults. Absolute FRC differences (Figure 4a), relative FRC differences (Figure 4b), mean difference and upper and lower limits of agreement (lines) are plotted against mean FRC of both tests. Mean difference is −15.4 mL (−1.1%), upper and lower limits of agreement are 793.3 mL (23.2%) and −824.1 mL (−25.4%).
Figure 5.
In vivo variability of lung clearance index.
Bland Altman plot [16] of lung clearance index (LCI) measured on two study days within three weeks in 30 healthy adults. Absolute LCI differences (Figure 5a), relative LCI differences (Figure 5b), mean difference and upper and lower limits of agreement (lines) are plotted against mean LCI of both study days. Mean difference is −0.13 (−1.8%), upper and lower limits of agreement are 0.55 (7.6%) and −0.81 (−11.2%).