Background Persistent pulmonary dysfunction is common after COVID-19, yet traditional assessments using carbon monoxide diffusing capacity (DLCO) alone may miss alveolar-capillary impairment. Objective To determine whether combining nitric oxide (DLNO 5s) and carbon monoxide (DLCO 5s) diffusing capacities enhances detection of post-COVID-19 lung impairment and whether summed z-scores outperform individual measures in classifying affected individuals. Design and methods We conducted an individual participant data meta-analysis using hierarchical mixed-effects modelling. The dataset included 572 COVID-19 survivors and 72 matched controls from six European centres. Lung function metrics - including spirometry, total lung capacity, DLNO 5s and DLCO 5s - were standardised into z-scores. Logistic models were compared using Bayesian Information Criterion and Leave-One-Out Information Criterion. Classification accuracy was assessed with Matthews Correlation Coefficient (MCC) and net reclassification improvement (NRI). Principal Component Analysis examined score structures, and dyspnoea severity was correlated with z-scores. Assessments were conducted 32-575 days post-infection (median=130 days). Results The number of days between SARS-CoV-2 diagnosis and testing did not affect any of the measured z-scores. Summed DLNO 5s + DLCO 5s z-scores consistently outperformed individual metrics. The combined model improved MCC by 0.06 (95% CI 0.01 to 0.11) and NRI by 37% (95% CI 13 to 62%) over DLCO 5s alone. The top model summed DLNO 5s + DLCO 5s model explained 10% of fixed and 59% of random variance. DLCO 5s alone failed to identify reduced membrane diffusion in approximately 16% of cases. Dyspnoea severity was significantly associated with all diffusion indices (p<0.001), though combined scores showed no stronger correlation than single predictors. Conclusion Summed DLNO 5s + DLCO 5s z-scores enhance classification of post-COVID-19 pulmonary impairment beyond DLCO 5s alone. The NO-CO double diffusion approach offers improved diagnostic discrimination between previously infected individuals and controls and aligns with symptom severity. These findings support broader clinical integration of combined diffusion metrics in post-COVID assessment.
Enhanced detection of patients with previous COVID-19: superiority of the double diffusion technique / G.S. Zavorsky, G. Barisione, T. Gille, R.W. Dal Negro, M. Núñez-Fernández, L. Seccombe, G. Imeri, F. Di Marco, J. Mortensen, E. Salvioni, P. Agostoni, V. Brusasco. - In: BMJ OPEN RESPIRATORY RESEARCH. - ISSN 2052-4439. - 12:1(2025 Aug 25), pp. e002561.1-e002561.10. [10.1136/bmjresp-2024-002561]
Enhanced detection of patients with previous COVID-19: superiority of the double diffusion technique
F. Di Marco;P. AgostoniPenultimo
;
2025
Abstract
Background Persistent pulmonary dysfunction is common after COVID-19, yet traditional assessments using carbon monoxide diffusing capacity (DLCO) alone may miss alveolar-capillary impairment. Objective To determine whether combining nitric oxide (DLNO 5s) and carbon monoxide (DLCO 5s) diffusing capacities enhances detection of post-COVID-19 lung impairment and whether summed z-scores outperform individual measures in classifying affected individuals. Design and methods We conducted an individual participant data meta-analysis using hierarchical mixed-effects modelling. The dataset included 572 COVID-19 survivors and 72 matched controls from six European centres. Lung function metrics - including spirometry, total lung capacity, DLNO 5s and DLCO 5s - were standardised into z-scores. Logistic models were compared using Bayesian Information Criterion and Leave-One-Out Information Criterion. Classification accuracy was assessed with Matthews Correlation Coefficient (MCC) and net reclassification improvement (NRI). Principal Component Analysis examined score structures, and dyspnoea severity was correlated with z-scores. Assessments were conducted 32-575 days post-infection (median=130 days). Results The number of days between SARS-CoV-2 diagnosis and testing did not affect any of the measured z-scores. Summed DLNO 5s + DLCO 5s z-scores consistently outperformed individual metrics. The combined model improved MCC by 0.06 (95% CI 0.01 to 0.11) and NRI by 37% (95% CI 13 to 62%) over DLCO 5s alone. The top model summed DLNO 5s + DLCO 5s model explained 10% of fixed and 59% of random variance. DLCO 5s alone failed to identify reduced membrane diffusion in approximately 16% of cases. Dyspnoea severity was significantly associated with all diffusion indices (p<0.001), though combined scores showed no stronger correlation than single predictors. Conclusion Summed DLNO 5s + DLCO 5s z-scores enhance classification of post-COVID-19 pulmonary impairment beyond DLCO 5s alone. The NO-CO double diffusion approach offers improved diagnostic discrimination between previously infected individuals and controls and aligns with symptom severity. These findings support broader clinical integration of combined diffusion metrics in post-COVID assessment.| File | Dimensione | Formato | |
|---|---|---|---|
|
e002561.full.pdf
accesso aperto
Tipologia:
Publisher's version/PDF
Licenza:
Creative commons
Dimensione
4.15 MB
Formato
Adobe PDF
|
4.15 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.




