Background Breathing pattern cannot be controlled in small children, so multiple breath washout SnIII analysis has to exclude inadequate volume breaths.
Aim To compare an existing mathematical breath exclusion algorithm with a physiological method.
Methods School age children with CF (30) and controls (30) performed SF6MBW with mass spectrometer, with uncontrolled tidal breathing. Two different breath exclusion methods were compared, with exclusion based on:
1) Expired tidal volume (VT) deviating by >25% of the median VT1
2) VT <3 Langley dead space2 volume or 90% bigger than the median VT
Runs with >33% excluded breaths were removed. Volume corrected Scond was calculated from subjects with 3 valid runs.
Results Far fewer subjects were excluded by the physiological Langley method, than by the mathematical method (Table). The mean and SD for Scond was identical by both methods, implying that the mathematical algorithm excludes valid data.
Conclusion A physiological approach to data cleaning prior to SnIII analysis allows retention of data that would be inappropriately excluded mathematically.