Chest computed tomography findings do not influence the decision of pneumologists regarding the diagnosis and management of pulmonary long coronavirus disease: a single center retrospective study

Introduction

Recent years have shown that not only the acute complications, but also long-term health consequences of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can be devastating. This includes a variety of physical, cognitive, and psychological symptoms that negatively impact daily functioning and quality of life. The term “long coronavirus disease (COVID)”, initially coined in social media, is defined by the United Kingdom National Institute for Health and Care Excellence (NICE) guideline recommendation as health complaints that persist or re-emerge 4 weeks beyond the acute illness phase of SARS-CoV-2 infection (1).

According to an Umbrella review, the frequency of long COVID in studies of adults without hospitalization varied from 7.5% to 41% (2). Among adults hospitalized for COVID-19 disease, long-term health sequelae were reported in 37.6% (2). Due to the magnitude of patients affected, this is a major problem not only for healthcare providers but also for economy. Currently, specialized outpatient clinics are being set up in Germany to meet the needs of the long COVID clientele. In this regard, current national and international guidelines recommend algorithms for the management of this syndrome. These include a sequence that begins with primary care, continues through specialized inpatient care or acute inpatient hospitalization, and, if long COVID syndrome persists, recommend remediation or rehabilitation (3,4). As for pulmonary long COVID chest imaging is recommended for a subgroup of patients (3), however, there is a lack of scientific evidence on the timing of imaging and the adequate modality. Do the benefits outweigh the effort and radiation exposure especially associated with computed tomography (CT)? What impact does imaging have on the diagnosis and the downstream management of patients?

To our knowledge a prior study has not been conducted to evaluate the impact of CT on the work-up of patients suspected to suffer from pulmonary long COVID. We closed this gap with the present study by assessing the effect of CT on the diagnostic decision regarding the measures taken for work-up, regarding the final diagnosis, the confidence in the diagnosis, and on the downstream management of patients. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-433/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by institutional ethics committee of the University of Regensburg (No. 20-1784-104) and individual consent for this retrospective analysis was waived.

Patients

All patients presented in a dedicated long COVID outpatient clinic of a secondary care hospital specialized on lung diseases (Caritas Hospital St. Maria Donaustauf) between April 2020 and August 2021. The referring physicians were either a primary care physician or a pulmonologist in line with algorithms recommended by the guidelines. The inclusion criteria were age ≥18 years, suspicion for long COVID syndrome [as defined in the NICE guideline (1)] of pulmonary origin (e.g., chronic cough, dyspnea) and availability of a chest CT acquired during work-up. Patient data (gender, age, eventual lung diseases existing before the acute COVID infection) and characteristics of the acute COVID-infection [like need for hospital admission, acute respiratory distress syndrome (ARDS), need for ventilation and/or extracorporeal membrane oxygenation (ECMO)] were extracted from electronic patient records or from letters provided by the referring physicist.

Patient work-up

Work-up in the long COVID outpatient clinic was standardized and consisted of clinical examination including auscultation, blood gas analysis, laboratory diagnostics, pulmonary function test and 6-minute walking test. CT was performed when one of the following was altered: lung function and/or diffusion capacity and/or auscultation. CT was done without contrast unless pulmonary artery embolism (PAE) was suspected. PAE was suspected in patients for whom at least one of the following was true: dyspnea and/or thoracic pain, signs of right heart strain, D-dimer increased in the three-digit range (normal range 45–500 ng/mL) without any other recognizable cause or diffusion impairment [diffusing capacity of the lung for carbon monoxide (DCLO) <80%] together with at least one of the first three parameters mentioned.

CT image acquisition

The CT protocol in the study hospital was as follows: acquisition on a 128-slice Somatom Definition AS by Siemens Healthcare, Erlangen, Germany, automatic tube voltage selection with a reference tube voltage of 120 kV, automatic tube current modulation with the reference mAs between 20–352. Multiplanar reformations (MPR) were reconstructed in the axial plane with a slice thickness of 1 mm in lung kernel and soft tissue kernel. Additional sagittal and coronal MPRs were reconstructed with a slice thickness of 1 mm in lung kernel. In total, 51% (21/41) of the scans were contrast enhanced (CE) and 49% (20/41) unenhanced (UE). Iohexal (Accupaque 350, GE Healthcare, Chicago, Illinois, USA) was administered in a weight-adapted dose, but not more than 70 mL. Flow rate was 3 or 4 mL/s.

Nine of the 41 CT scans were performed at external hospitals. Two of them were CE and seven of them without contrast. The slice thickness varied between 1 and 4 mm in the lung kernel (four times 1 mm, one time 1.25 mm, one time 1.5 mm, two times 3 mm, one time 4 mm) and between 3 and 5 mm in the soft tissue kernel (five times 3 mm, two times 4 mm, two times 5 mm). All pictures were sent to the picture archiving and communication system (PACS, Syngo Imaging, Siemens, Erlangen, Germany).

CT image analysis

All CT images (acquired both in house and outside) were evaluated by a dedicated chest radiologist with 23 years of experience. Fleischner criteria were applied (5). All available prior imaging (if performed) including imaging of the acute phase was considered for the final report. The final report contained a description of all pathologies, a comment on the disease course and a statement about the most likely diagnosis [in terms of scarring, inflammation, interstitial lung disease (ILD)…].

Analysis of patient data

All cases were evaluated independently by three senior pneumologists (I.D., D.S, M.M.) with 22, 14 and 14 years of experience, respectively. Pneumologist 1 and 3 worked in the secondary care hospital specialized on lung diseases, pneumologist 2 worked in the Department of Pneumology of the Affiliated University Hospital. All pneumologists evaluated the patients in two rounds. The time in between the two rounds was at least 2 months in order to avoid a recall bias. In the first round, the pneumologists were unaware of the results of the CT scan performed in the context of long COVID work-up. In the second round, the CT report was revealed. The remainder of information was identical in the two rounds. The three evaluators were given the patient’s records. The evaluators were blinded to personal data such as name and age. All clinical and laboratory data as well as results of prior imaging could be reviewed. For every patient the following identical parameters were queried in both of the two runs: diagnosis of pulmonary long COVID (yes/no), confidence of the diagnosis on a scale from 0 to 3 (0: very uncertain, 1: uncertain, 2: certain, 3: very certain), need for: bronchoalveolar lavage (BAL) (yes/no), TBB (yes/no), cryobiopsy (yes/no), video-assisted thoracoscopy (VATS) (yes/no), ergospirometry (yes/no), ventilation/perfusion scintigraphy (yes/no), follow-up appointment (yes/no), rehabilitation (yes/no).

Statistical analysis

Age is presented as mean [standard deviation (SD)] and all categorical variables as absolute and relative frequencies. Continuous data are reported as mean (SD), ordinal data as median (range), and categorical data as absolute frequencies (percentage). SD or 95% confidence intervals (CIs) are reported where applicable. The statistical analysis did not require controlling for the family-wise error rate (FWER). Cohen’s kappa coefficient was used to quantitatively assess interrater reliability. Strength of agreement was measured according to Landis and Koch: <0= poor, 0.01–0.2= slight, 0.21–0.4= fair, 0.41–0.6= moderate, 0.61–0.8= substantial and 0.81–1= almost perfect (6). Paired sample t-tests were used to compare the confidence of diagnosis with and without knowledge of the CT report. Differences of decisions regarding diagnosis of COVID-associated pulmonary disease (PD), further diagnostic measures, and all parameters of downstream management between the two rounds were analyzed by using the McNemar-Test. A P value ≤0.05 was considered statistically significant.

All analyses were performed using R, version 3.6.1 (The R Foundation for Statistical Computing, Vienna, Austria).

Results

Overall, 628 patients presented in the long COVID outpatient clinic during the inclusion period. Five hundred eighty-seven patients had to be excluded because they did not receive a chest CT. Forty-one patients met the inclusion criteria of the study. The study population consisted of 24 males (59%) and 17 females (41%) patients, aged from 21 to 72 years (mean 55 years, SD 9).

Before and during acute COVID-19 infection

Twelve percent (5/41) of patients had a pre-existing ILD. These included two cases of hypersensitivity pneumonitis, one case of sarcoidosis and two cases of unclassifiable ILD. Apart from asthma in four more cases, no other lung diseases were present. In the acute phase of the infection, 17% of patients (7/41) needed mechanical ventilation, 7% (3/41) needed ECMO treatment and 15% (6/41) suffered from ARDS. In 32% (13/41) of patients a chest X-ray of the acute phase was available. In 49% (20/41) of cases, a chest CT scan of the acute phase was available. Out of these 20 scans, PAE was seen in 20% (4/20) and ruled out in 50% (10/20). The remaining 30% of scans (6/20) had been performed UE. Please see Table 1.

Diagnostic decisions depending on knowledge of CT result

Diagnosis of COVID-associated PD

First round (without knowledge of the findings of CT performed during work up for long COVID)

In an average of 24% (10/41) of patients, a diagnosis of pulmonary long COVID was made. For the individual assessments of the pneumologists please see Table 2.

Second round (with knowledge of the findings of CT performed during work up of long COVID)

Diagnosis of pulmonary long COVID was made in an average of 27% (11/41) of patients. For the individual assessments of the pneumologists please see Table 2.

The decision of all three raters taken together regarding COVID-associated PD did not change significantly without and with knowledge of the CT report (P=0.69). Please see Table 3.

Confidence of diagnosis

Between the two rounds the mean (95% CI) confidence of diagnosis increased from 1.9 (1.8, 2.1) to 2.9 (2.8, 3) in Rater 1 (P<0.001) and from 2.2 (2, 2.5) to 2.7 (2.6, 2.9) in Rater 3 (P<0.001), see Table 2. Although not evident from mean and CI, the frequency distribution also shows a significant increase in confidence for Rater 2 between rounds 1 and 2 (P=0.02), see Table 2. Overall, the mean confidence of diagnosis increased from 1.9 (1.8, 2.1) to 2.6 (2.4, 2.7) (P<0.001), see Table 3.

Further diagnostic measures and downstream management depending on knowledge of CT result

The decision of Rater 1 regarding further diagnostic measures and downstream management did not change significantly. Rater 2 ordered significantly more ergospirometries (P=0.001) and significantly less follow-ups (P=0.05) after knowledge of the CT findings. Rater 3 was significantly less likely to initiate rehabilitation with knowledge of the CT findings (P<0.001). Please see Figure 1 and Tables 2,4.

Figure 1 Decisions of pneumologists regarding diagnostic measures and downstream management without (round 1) and with (round 2) knowledge of CT findings. BAL, bronchoalveolar lavage; CT, computed tomography; CT CE, CT contrast enhanced; CT UE, CT unenhanced; TBB, transbronchial biopsy; VATS, video-assisted thoracoscopy; V/P scintigraphy, ventilation perfusion scintigraphy.

Interobserver variability

Without knowledge of the CT findings, pneumologist 1 and 2 showed a slight strength of agreement for diagnosis of pulmonary long COVID (κ=0.19), pneumologist 1 and 3 showed a fair strength of agreement (κ=0.31) and pneumologist 2 and 3 showed a moderate strength of agreement (κ=0.44). With knowledge of the CT findings, pneumologist 1 and 2 showed a slight strength of agreement (κ=0.11), pneumologist 1 and 3 showed a moderate strength of agreement (κ=0.54) and pneumologist 2 and 3 showed a slight strength of agreement (κ=0.16). Please see also Table 5.

Discussion

The number of patients complaining of prolonged symptoms regarding their physical health and wellbeing long after an acute infection SARS-CoV-2 is increasing (7,8). The term long COVID was initially coined in social media by Dr. Elisa Perego in May 2020 (9). It took no less than 7 months for it to be defined for the first time by the United Kingdom National Health Service (1), followed by the adoption by other institutions worldwide (10,11). The disease, which did not exist 3 years ago, currently affects approximately 1.9 million Britons according to National Health Service (NHS) (12).

Chest CT played and continues to play an important role in the work-up of COVID-19. A vast number of studies have been published on pulmonary abnormalities during and after acute infection with SARS-CoV-2. Several papers have reported on long term lung sequelae (13-20). Given the current situation, there is likely to be an increasing demand for imaging (21). However, to the best of our knowledge the role of CT in the assessment, diagnosis and management of pulmonary long COVID has not yet been explored. The aim of this retrospective study was to fill this gap.

A total of 41 patients suspected to suffer from pulmonary long COVID were independently analyzed by three pneumologists. For all patients a chest CT was performed as part of the work-up. In the first round, pneumologist were blinded to the results of CT, in the second round CT findings were revealed.

Diagnosis of pulmonary long COVID

Pulmonary long COVID was diagnosed in 24% of patients in the first round and 27% in the second round. Thus, the diagnosis was confirmed only in a minority of patients although all of them had demonstrated an impaired lung function test or a conspicuous auscultation. Knowledge of the CT results did not influence the pneumologists’ assessment. However, the confidence of diagnosis was significantly higher knowing the CT findings.

Interobserver variability was slight to moderate and did overall not reach clinical requirements. Of note, knowledge of CT findings did not improve the level of agreement. Obviously, clinicians seem to have difficulties to classify the clinical picture adequately. Pulmonary long COVID is a complex condition with a heterogeneous appearance and the existing guidelines seem not to be suited to allow for a robust classification.

Further diagnostic measures and downstream management

In the vast majority of cases, decision on further diagnostic measures and downstream management of patients was not altered by the knowledge of the CT findings. One rater wished for significantly more ergospirometries but significantly less follow-ups when provided with the CT findings. Another rater opted for significantly less rehabilitation measures. Only one rater wished for significantly less diagnostic measures in total, but the other two raters did not show a significant difference regarding their request of examinations. Overall, there also was no significant difference without and with knowledge of the CT report. Thus, CT might increase the wish for further functional assessment but help to avoid unnecessary downstream measures.

Interestingly a significant difference was found between the raters regarding the number of further diagnostic measures wished for. This difference persisted independently of CT findings. The pneumologist employed in the university hospital (Rater 2) ordered significantly more examinations than the other two raters who worked in the district hospital (specialized lung clinic). Patient handling seems to be influenced by the available resources. Also, Rater 1 ordered significantly less examinations than Rater 2 and Rater 3. Rater 1 was by far the most experienced physician, which may be the reason for this finding.

The study has limitations. According to the hospital specific standard operating procedure the indication for a CT in the work-up of suspected pulmonary long COVID was limited to patients with restriction on body plethysmography, diffusion impairment (DLCO <60%) or sclerosiphonia on auscultation. Thus, the cohort represented a subset of patients with a rather high probability for lung parenchyma pathologies. Even in this subgroup CT did not alter the diagnostic decisions. Hence, it is very unlikely that this would be the case for less severely affected patients.

The long term outcome of the patients was not registered. It is therefore not possible to evaluate if the pneumologists classified the patients correctly or if the CT findings could have predicted the outcome. However, the aim of the study was to identify the impact of CT findings on the diagnostic decisions regardless of the long term consequences.

Conclusions

In summary, our analysis revealed that chest CT did not influence the diagnostic decision for patients suspected to suffer from pulmonary long COVID. The confidence in the diagnosis, on the contrary, was improved by chest CT. Still, based on our results CT seems not to be justified in the work-up of long COVID when pulmonary function tests and auscultation are normal. Interobserver variability regarding the diagnosis of pulmonary long COVID was rather low and did not increase by chest CT. Obviously, the existing guidelines are not specific enough to allow for a clear classification of patients.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-433/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-433/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-433/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-433/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by institutional ethics committee of the University of Regensburg (No. 20-1784-104) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. National Institute for Health and Care Excellence: Clinical Guidelines. COVID-19 rapid guideline: managing the long-term effects of COVID-19. London: National Institute for Health and Care Excellence (NICE). Copyright © NICE 2020; 2020.
2. Nittas V, Gao M, West EA, et al. Long COVID Through a Public Health Lens: An Umbrella Review. Public Health Rev 2022;43:1604501. [Crossref] [PubMed]
3. Koczulla AR, Ankermann T, Behrends U, et al. S1-Leitlinie Long-/Post-COVID. Pneumologie 2022;76:855-907. [Crossref] [PubMed]
4. Clinical management of COVID-19: living guideline. World Health Organization; 2023.
5. Hansell DM, Bankier AA, MacMahon H, et al. Fleischner Society: glossary of terms for thoracic imaging. Radiology 2008;246:697-722. [Crossref] [PubMed]
6. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.
7. Davis HE, McCorkell L, Vogel JM, et al. Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol 2023;21:133-46. [Crossref] [PubMed]
8. Joseph G, Margalit I, Weiss-Ottolenghi Y, et al. Persistence of Long COVID Symptoms Two Years After SARS-CoV-2 Infection: A Prospective Longitudinal Cohort Study. Viruses 2024;16:1955. [Crossref] [PubMed]
9. Perego E. Twitter post.Available online: https://twitter.com/elisaperego78/status/1263172084055838721 (accessed August 23, 2025).
10. Rabady S, Altenberger J, Brose M, et al. Guideline S1: Long COVID: Diagnostics and treatment strategies. Wien Klin Wochenschr 2021;133:237-78. [Crossref] [PubMed]
11. Antoniou KM, Vasarmidi E, Russell AM, et al. European Respiratory Society statement on long COVID follow-up. Eur Respir J 2022;60:2102174. [Crossref] [PubMed]
12. Office for National Statistics (ONS). Prevalence of ongoing symptoms following coronavirus (COVID-19) infection in the UK: 30 March 2023. Statistical bulletin. Released 30 March 2023. Available online: https://www.ons.gov.uk/ (accessed August 23, 2025).
13. Han X, Fan Y, Alwalid O, et al. Six-month Follow-up Chest CT Findings after Severe COVID-19 Pneumonia. Radiology 2021;299:E177-86. [Crossref] [PubMed]
14. Sonnweber T, Sahanic S, Pizzini A, et al. Cardiopulmonary recovery after COVID-19: an observational prospective multicentre trial. Eur Respir J 2021;57:2003481. [Crossref] [PubMed]
15. Lee JH, Yim JJ, Park J. Pulmonary function and chest computed tomography abnormalities 6-12 months after recovery from COVID-19: a systematic review and meta-analysis. Respir Res 2022;23:233. [Crossref] [PubMed]
16. Li D, Liao X, Ma Z, et al. Clinical status of patients 1 year after hospital discharge following recovery from COVID-19: a prospective cohort study. Ann Intensive Care 2022;12:64. [Crossref] [PubMed]
17. Han X, Chen L, Fan Y, et al. Longitudinal Assessment of Chest CT Findings and Pulmonary Function after COVID-19 Infection. Radiology 2023;307:e222888. [Crossref] [PubMed]
18. Cha MJ, Solomon JJ, Lee JE, et al. Chronic Lung Injury after COVID-19 Pneumonia: Clinical, Radiologic, and Histopathologic Perspectives. Radiology 2024;310:e231643. [Crossref] [PubMed]
19. Bocchino M, Rea G, Capitelli L, et al. Chest CT Lung Abnormalities 1 Year after COVID-19: A Systematic Review and Meta-Analysis. Radiology 2023;308:e230535. [Crossref] [PubMed]
20. Fang X, Lv Y, Lv W, et al. CT-based Assessment at 6-Month Follow-up of COVID-19 Pneumonia patients in China. Sci Rep 2024;14:5028. [Crossref] [PubMed]
21. Alghamdi F, Owen R, Ashton REM, et al. Post-acute COVID syndrome (long COVID): What should radiographers know and the potential impact for imaging services. Radiography (Lond) 2022;28:S93-9. [Crossref] [PubMed]