Z -Scores or Fixed Cutoff Values to Interpret Oscillometry: Opportunities from the LEAD Study

This is the authors' abstract. We don't add key points for this paper.

In a large community-based, cross-sectional, epidemiological study conducted among 7,560 male and female subjects older than 18 years, Veneroni and colleagues (1) examined the performance of lung oscillometry (Resmon Pro Full, Restech Srl) and spirometry among subjects with self-reported respiratory symptoms and respiratory diagnosis as part of the Austrian LEAD (Lung, Heart, Social, Body) study.The authors reported that 20% of the participants had abnormal findings on oscillometry, compared with 13% with abnormal spirometry results.Respiratory symptoms and/or respiratory diagnoses were present in 28.7% of the study population, and the odds of abnormal findings on oscillometry in this subgroup were significantly higher than among those with abnormal spirometry results.Moreover, 17% of symptomatic/diseased subjects who had normal findings on spirometry showed abnormal oscillometry results.The oscillometry parameters used were respiratory resistance at 5 Hz (R5) (total, inspiratory, and expiratory), respiratory reactance at 5 Hz (X5) (total, inspiratory, and expiratory) and area of reactance (AX).R5 minus respiratory resistance at 20 Hz and resonant frequency parameters were not reported because of a lack of reference values.Abnormal findings on oscillometry were defined as Z-score .1.64(for R5 and AX) and ,21.64 (for X5) using the reference equation generated by Oostveen and colleagues in 2013 (2).The authors concluded that lung oscillometry has the potential to detect and evaluate respiratory impairment in the community, even among individuals with normal spirometry results.Interpreting spirometry necessitates the use of locally derived reference values because age, gender, ethnicity, height, and weight account for 70-90% of the variability.All spirometry indices are normally distributed, and the values of individual subjects are interpreted using Z-score (i.e., the deviation from the standard values).Generally, Z-score below (or above) 1.64 (i.e., 2 SDs) are considered to indicate abnormality.In contrast to spirometry, all studies have shown that oscillometry variables are not normally distributed (3), including that of Oostveen and colleagues (2).If the data are not normally distributed, then the Z-score distributions are also skewed.The defining cutoff values to differentiate between normal and abnormal used in the study (Z-score .1.64 for R5 and Ax and Z-score , 1.64 for X5) may therefore not be reliable.