Introduction

The COVID‑19 pandemic encompassed the period from December 2019 to May 2023. It developed due to SARS‑CoV‑2 outbreak, causing more than 778 691 725 disease cases and 7 102 530 deaths worldwide.¹ The outbreak resulted in an overwhelming number of patients experiencing persistent respiratory symptoms long after the initial phase of the disease. The prolonged consequences of viral pneumonia included persistent lung function impairment and radiological changes, such as ground‑glass opacities and signs of lung fibrosis, indicating interstitial lung disease (ILD).² Apart from persistent radiological findings, chronic lung damage, characterized by impaired lung function, especially decreased transfer factor of the lungs for carbon monoxide (TL_CO), may also last for months after the initial infection.^2,3 Furthermore, COVID‑19 survivors experienced cardiological and neurological complications. The former were manifested by arrhythmias, acute coronary syndrome, and myocarditis. The latter, by anosmia, headaches, vertigo, seizures, and strokes.⁴ Many patients continued to experience a wide range of symptoms for weeks following their discharge from hospital. According to recent literature, the term “long COVID” may be applied when referring to the presence of a chronic condition affecting 1 or many organs, lasting more than 3 months after SARS‑CoV‑2 infection.⁵ In the case of pulmonary outcomes, development of lung fibrosis after recovering from SARS‑CoV‑2 infection may lead to persistent lung damage.⁶ To seek effective, noninvasive methods for monitoring and evaluating the pulmonary sequelae of COVID‑19, lung ultrasound (LUS) has emerged as a potential diagnostic tool.³ While chest computed tomography (CT) remains the gold standard for radiological imaging, LUS has been extensively applied in the acute phase of the disease, offering a bedside imaging opportunity. Its utility in the follow‑up assessment of the disease course continues to be underexplored.^3,7-10

This study aimed to assess the correlation between LUS findings and key pulmonary function test (PFT) results in patients with persistent clinical symptoms after COVID‑19 pneumonia, and to determine the utility of LUS as a screening tool for detecting pulmonary impairment in COVID‑19 survivors.

Patients and methods

An observational, prospective study was conducted in a single pulmonology department between January 1, 2021 and September 30, 2022. It included patients who had previously been diagnosed with COVID‑19 pneumonia and continued to experience respiratory symptoms postrecovery. The study protocol was approved by the Bioethics Committee at the National Institute of Tuberculosis and Lung Diseases in Warsaw (KB‑92/2020). The participants provided written informed consent prior to inclusion. All patients met the following inclusion criteria: 1) age over 18 years; 2) confirmed SARS‑CoV‑2 infection (positive nasopharyngeal swab in antigen testing or reverse transcription polymerase chain reaction testing, or presence of SARS‑CoV‑2–specific immunoglobulin (Ig) M or IgG antibodies in patients not vaccinated against SARS‑CoV‑2 infection); 3) residual lung abnormalities after viral pneumonia on radiological imaging, and 4) chronic respiratory symptoms. Exclusion criteria comprised pre‑existing chronic lung diseases unrelated to COVID‑19 and inability to perform PFTs or undergo a LUS examination. The LUS and PFTs were performed less than 21 days apart.

Results

Discussion

Since the outbreak of the COVID‑19 pandemic, various observational studies have been published, expanding our knowledge of late pulmonary sequelae following infection. Post–COVID‑19 patients showed altered respiratory function, with a decrease in TL_CO reported in up to 40% of the cases.¹³ Lung function alterations and radiological changes were more often observed in the groups with the most severe disease course.¹⁴ In recent publications, LUS showed a good correlation with CT scans for identifying persistent radiological changes.^3,7-9,15 A detailed description of the diagnostic performance of LUS relative to CT in this cohort has been published previously, demonstrating a strong correlation between LUS scores and CT‑derived quantification of the lung area affected by ground‑glass opacities, calculated by artificial intelligence (R = 0.702; P <⁠0.05).¹⁰ Over half of our study group presented with pulmonary function impairment (54.5%). A significant, moderate inverse correlation between the LUS score and key pulmonary function parameters (TL_CO, FVC, and FEV1) was observed. Moreover, a weak inverse correlation with TLC was noted. These findings may highlight the potential role of LUS in evaluating post–COVID‑19 patients, particularly those with persistent symptoms, suggesting that a higher degree of lung abnormalities observed and scored on ultrasound is associated with more severe lung function impairment. We also aimed to evaluate an optimal LUS score cutoff value for identifying persistent lung damage. A threshold of 8 points showed the highest diagnostic accuracy for detecting a TL_CO below 80% of predicted normal values. At this threshold, the sensitivity was 78.6%, specificity was 62.8%, and the overall accuracy reached 71.7%, with a Youden index of 0.414. These findings may indicate the potential of LUS as a screening tool for pulmonary diffusion impairment in post–COVID‑19 patients. The balance between sensitivity and specificity is particularly important in clinical settings where access to full PFTs is limited. These results are similar to those reported by Clofent et al,⁹ who identified a TL_CO decrease in 66.5% of their study population and a moderate diagnostic yield of LUS when assessing interstitial lung involvement, reinforcing its utility as a diagnostic tool for both structural and functional assessments. While LUS cannot replace TL_CO testing, it can serve as an efficient screening tool for identifying patients who would benefit from further functional or radiological assessment. These findings are in line with those of other emerging studies. Tana et al¹⁶ found that the LUS score was useful for predicting clinical deterioration and mortality in patients with COVID‑19. On the other hand, in a multicenter observational study by Hernandez et al,¹⁷ LUS showed a high negative diagnostic value, indicating that it could be used to rule out pulmonary sequelae after COVID‑19 pneumonia in the case of a normal image. Studies suggest that some radiological lesions tend to gradually resolve over time.⁷ However, individuals with fibrotic‑like changes experience more respiratory symptoms and have a significantly lower TL_CO. Therefore, these changes might indicate irreversible pulmonary sequelae.¹⁸ The possibility of distinguishing between reversible inflammation and permanent fibrosis using LUS alone is yet to be explored, particularly in the absence of consecutive follow‑up or confirmatory radiological imaging. The British Thoracic Society guidelines¹⁹ recommend performing PFTs at 3 months postdischarge, especially in the patients suspected of ILD following COVID‑19. LUS may be integrated into the follow‑up assessment at 3 months to screen for ILD.³ Notably, in this study repeated measurements were performed in a subgroup of patients, adding to the evaluation of lung recovery. The consistent pattern across our study and in complementary studies mentioned above suggests a reproducible association between abnormalities identified on LUS and impaired PFT results, particularly TL_CO. This is clinically important, as TL_CO is often the most sensitive PFT parameter for early ILD diagnosis, and may remain abnormal even when spirometry appears within normal limits.

Regarding clinical practice, this correlation supports the integration of LUS into follow‑up protocols after initial assessment of COVID‑19 survivors, especially when access to repeated high‑resolution CT or full PFTs is limited. Taking into consideration the portability, affordability, and favorable safety profile of LUS (no radiation exposure), it may be a good pragmatic tool for screening and longitudinal follow‑up.