Longer hospital stay was observed to be the only independent predictor of CS-DPLA six months after severe COVID-19. Serum TGF-β should be evaluated further as a biomarker.

The study participants included patients who recovered after acute severe COVID-19 and presented with CS-DPLA at two or six month follow up and control group (without CS-DPLA). Adults volunteers without any acute illness, chronic respiratory illness and without a history of severe COVID-19 were included as healthy controls for the biomarker study. The CS-DPLA was identified as a multidimensional entity involving clinical, radiological and physiological pulmonary abnormalities. The primary exposure was the neutrophil-lymphocyte ratio (NLR). Recorded confounders included age, sex, peak lactate dehydrogenase (LDH), advanced respiratory support (ARS), length of hospital stay (LOS) and others; associations were analyzed using logistic regression. The baseline serum levels of surfactant protein D, cancer antigen 15-3 and transforming growth factor-β (TGF-β) were also compared among cases, controls and healthy volunteers.

We identified 91/160 (56.9%) and 42/144 (29.2%) participants with CS-DPLA at two and six months, respectively. Univariate analyses revealed associations of NLR, peak LDH, ARS and LOS with CS-DPLA at two months and of NLR and LOS at six months. The NLR was not independently associated with CS-DPLA at either visit. Only LOS independently predicted CS-DPLA at two months [adjusted odds ratios (aOR) (95% confidence interval [CI]), 1.16 (1.07-1.25); P<0.001] and six months [aOR (95% CI) and 1.07 (1.01-1.12); P=0.01]. Participants with CS-DPLA at six months had higher baseline serum TGF-β levels than healthy volunteers. Interpretation and conclusions: Longer hospital stay was observed to be the only independent predictor of CS-DPLA six months after severe COVID-19. Serum TGF-β should be evaluated further as a biomarker.