Characterisation of leaks during multiple breath washout and their influence on test results

Background: Multiple breath washout (MBW) is a sensitive test to assess ventilation inhomogeneity (VI). Based on ERS/ATS guidelines traces need to be discarded if leaks occur. However, it is unclear if this is needed for all types of leaks. Objectives: To systematically characterize different leak types in N 2 MBW and assess their influence on lung clearance index (LCI) and functional residual capacity (FRC). Methods: Experimental study using existing data of 5 children with CF and 5 controls. Reanalysis of traces was done after modelling different leak types. Leaks were categorized based on location, in- or expiration, duration, stage of washout and size. We verified theoretical results by performing real-time N2MBW measurements with artificially introduced leaks. Results: FRC and LCI are both significantly influenced by all leak characteristics. A precapillary expiratory continuous leak of e.g. 20% after ¼ of full washout time changed LCI from mean 7.0 to 6.3 (p Conclusions: The influence of leaks on N 2 MBW outcome is very heterogenous. Depending on the respective purpose not all measurements with suspected leaks need to be excluded. There are simple ways to recognize different types of leaks and distinguish them from increased VI.