Quality control of spirometry in children: Can ERS/ATS criteria replace visual inspection?

Expert9s visual rating is an integral part of quality control for spirometry but is not always done. In our multi-centre reference study (LuNoKiD) we investigated whether quality criteria based on back-extrapolated-volume (BeV), reproducibility and forced-expiratory-time (FET) are feasible and can replace visual inspection. Caucasian children 4 to 18 years old were recruited. Spirometry was carried out according to international standards under field conditions. Experts rated visual acceptability of each curve (9 per individual). We tested feasibility, sensitivity and specificity of the currently recommended quality control criteria, and determined optimal cut-offs and combinations of the three computer-based quality measures using visual acceptability as gold standard. A total of 3133 of 5104 investigated subjects were healthy and included. 72% of these had visually acceptable flow-volume curves. In this group, 95% met ATS/ERS recommendations for BeV, 90% for reproducibility and 47% for FET. Best correspondence with visual acceptability was reached when choosing the cut-offs 4.6% for BeV (in percent FVC) and 1.8 sec for FET. The best combination of the three measures showed good overall concordance with visual acceptability but still a specificity of 56% only. Our study supports the hypothesis that - in contrast to BeV and reproducibility - ATS/ERS-recommended cut-offs for FET are not feasible under field conditions. The low specificity of the optimal combination of the three quality measures further demonstrates that these cannot replace visual control.