Natural Variability of Clinically Measured Lung Clearance Index in Children with Cystic Fibrosis

Abstract

<b>Rationale:</b> The lung clearance index (LCI) is increasingly used in clinical surveillance of patients with cystic fibrosis (CF). However, information on the natural variability of LCI are sparse and the minimal clinically important difference (MCID) as threshold to guide clinical decisions remains unknown. <b>Objective:</b> To describe natural variability of clinical LCI in children with CF at stability on a population and individual level and define potential MCID thresholds. <b>Methods:</b> Children with CF aged 4 - 18 years performed LCI measurements every three months as part of routine clinical surveillance during 2011-2018 in two centers. Natural variability during periods of clinical stability was calculated on a population level using mixed-effects models. On an individual level, limits of relative changes in LCI for each patient were calculated separately. To derive MCID, we compared different cut-offs indicating LCI changes higher than natural variability. <b>Results:</b> Repeated LCI measurements of acceptable quality (N= 741) were available in 97 patients with CF. On a population level, variability of LCI expressed as coefficient of variation (CV%) was 8.8%. Upper limit of normal (ULN) for relative changes in LCI between visits was 23% (95% quantile), but individual limits were heterogeneously distributed (range 2.5% to 49.1%). <b>Conclusion:</b> We report natural variability of LCI in the clinical setting and suggest potential MCID thresholds based on the limits of variability on a population and individual level. The large inter-individual variability suggests that applying individual limits might be an appropriate approach to define MCID for routine clinical surveillance.