The concordance of virtual coached spirometry in a COPD clinical trial

<bold>Introduction:</bold> Spirometry assessment of airflow limitation is a key outcome in respiratory studies. Remote coached assessment may offer a convenient, flexible alternative to patients considering participation in research. We assessed the concordance of coached spirometry performed face-to-face (F2F) vs virtually within a COPD clinical trial (<ext-link>NCT05492877</ext-link>). <bold>Methods:</bold> F2F and virtual coached spirometry were performed at baseline and 3 months. Spirometry quality was assessed by AI algorithm (ArtiQ.QC, Belgium) to ATS/ERS 2019 standards. Sessions obtaining A or B gradings were included. Concordance was assessed for FEV1 and FVC by intraclass correlation coefficient (ICC). <bold>Results:</bold> 381 participants were randomised. At baseline 226 participants had paired spirometry of sufficient quality. The mean (standard deviation [SD]) FEV1 was 1.33L (0.56L), 47.3% (17.0%) predicted. The mean (SD) FVC was 2.93L (0.87L), 81.3% (18.5%) predicted. The SD of the pairwise difference was 0.110L for FEV1 with an ICC of 0.98 and 0.280L for FVC with an ICC of 0.95. At 3 months 192 participants had paired spirometry. The mean variance was 0.117L for FEV1 with an ICC of 0.98 and 0.264L for FVC with an ICC of 0.96. The most common reason for failure to obtain a virtual result was connectivity. <fig><object-id>erj;66/suppl_69/PA6269/F1</object-id><object-id>F1</object-id><object-id>F1</object-id><graphic></graphic></fig> <bold>Conclusion:</bold> The concordance between virtual and F2F coached spirometry was high. Remote coached spirometry may provide a viable, more patient centric approach in clinical trials.