Introduction and objectives Multiple breath inert gas washout (MBW) is a technique for quantifying ventilation heterogeneity (VH). Commonly-used indices of VH such as lung clearance index (LCI) may be biased by alterations in functional residual capacity, tidal volume and anatomical dead space. We hypothesised that a novel marker of VH, exponential decay index (EDI) would exhibit more favourable measurement properties than established indices.
Methods Seventy-four patients with asthma and eighteen healthy control subjects were recruited. MBW was performed in triplicate in each subject using a modified Innocor gas analyser (Innovision A/S, Odense, Denmark). Analysis of the washout curve yielded a number of VH parameters including LCI, mixing ratio (MiR), moment ratio (MoR) and the novel parameter EDI. EDI was calculated by fitting the washout data to both one-phase and two-phase exponential decay models and deriving the difference in goodness of fit (R2) between the two models. Phase III slope analysis yielded values for Scond, Sacin and first breath SnIII (FB SnIII). Repeatability was assessed using intraclass correlation coefficient (ICC), and ability to discriminate between health and disease was determined using receiver operating characteristic (ROC) curves. Robustness to variations in FRC, Vt and Vd was determined by calculating the ratio of signal (absolute difference between healthy and asthma groups) to noise (standard deviation of calculated ‘ideal’ values for each subject, assuming perfect gas mixing).
Results The most repeatable parameters were LCI (ICC = 0.905), MoR (ICC = 0.869), FB SnIII (ICC = 0.867) and Sacin (ICC = 0.846). The most discriminatory parameters were FB SnIII, Sacin and EDI, with areas under the ROC curve of 0.728, 0.707 and 0.691 respectively. Signal-to-noise ratios for LCI, MiR and MoR were 1.726, 3.911 and 0.679 respectively, thus accounting for the poor discriminatory ability of these markers.
Conclusion FB SnIII and Sacin appear to be the most favourable markers of VH. EDI is the most discriminatory parameter that does not rely on phase III slope analysis. Since FB SnIII may be derived from a single-breath manoeuvre, this parameter warrants further investigation in other disease groups, and normative values should be derived in large healthy populations.