Pulmonary diffusing capacity and dyspnoea following COVID-19: Insights from multicentre datasets

Pulmonary complications remain a significant challenge for COVID-19 survivors, necessitating advanced diagnostic approaches for long-term assessment. We present a curated, open-access dataset of pulmonary function measurements-including nitric oxide (DLNO) and carbon monoxide (DLCO) diffusing capacities-in 572 post-COVID-19 patients and 72 healthy controls (filtered from an original cohort of 726 survivors and 126 controls). Collected across eight international centres, the data include demographics, spirometry, lung volumes, and 5-6 s single-breath DLNO_5s, DLCO_5s, and alveolar volume (VA_5s). Missing values for total lung capacity were imputed, and low-quality or system-specific (Hyp'Air Compact) measurements were excluded in the filtered dataset. A third subset (333 patients, 54 controls) links these measurements to dyspnoea severity (mMRC scale) for correlation and proportional odds analyses. This resource underpins predictive modeling of post-COVID pulmonary impairment via summed z-scores (DLNO + DLCO) and aims to accelerate validation of NO-CO diagnostics. The freely accessible datasets are provided in both SPSS (.sav) and .csv formats at the Mendeley Data Cloud-based repository and includes nominal, ordinal, and scalar data.