Degree of lung involvement on chest CT and prolonged corticosteroid use in critically ill patients with COVID-19: an observational exploratory study

Introduction

Chest computed tomography (CT) is the best imaging technique to assess the degree of lung involvement secondary to coronavirus disease 2019 (COVID-19) (1). Typical radiological findings include consolidations, peripheral ground-glass opacities, crazy-paving, and the reversed halo sign (2). The volume of pathological pulmonary opacities is associated with the degree of inflammation and risk of progressing to more severe disease (3). A previous study showed that pulmonary opacities exceeding 60% of lung volume were associated with a 19-fold increased risk of requiring mechanical ventilation (4).

In addition to the risk of acute respiratory distress syndrome (5), COVID-19 pneumonia can lead to late complications, such as pulmonary fibrosis (6), characterized by unsuccessful repair of the alveolar epithelium, persistence of fibroblasts, excessive deposition of collagen and other extracellular matrix components, and destruction of normal pulmonary architecture (7,8). A meta-analysis found that 44% of COVID-19 survivors developed pulmonary fibrosis to some extent, and identified the degree of lung involvement on chest CT as one of the main risk factors for post-COVID-19 pulmonary fibrosis (9).

In a meta-analysis of randomized studies evaluating interventions to manage COVID-19, using corticosteroids during the early phase was the only intervention with at least moderate certainty of reducing mortality and the need for mechanical ventilation compared to standard treatments (10). While corticosteroid therapy in the early phase of COVID-19 is well established (11,12), another study from our group suggested that prolonged corticosteroid use beyond 10 days is associated with increased mortality (13). The rationale for prolonged courses of corticosteroids in patients with COVID-19 is the control of exacerbated inflammatory responses, associated with the development of pulmonary fibrosis (14,15). Consequently, patients with greater pulmonary inflammatory involvement would theoretically derive greater benefit from prolonged therapy. This group primarily includes patients with extensive degree of radiological involvement and high volumes of pulmonary opacities identified by chest CT (4).

Although several studies have already investigated the association between corticosteroid treatment duration and pulmonary abnormalities in outpatients (16,17), the relationship between the degree of lung involvement observed in chest CT and the duration of corticosteroid treatment among COVID-19 patients remains unexplored. Thus, the main objective of this study is to evaluate the association between the duration of corticosteroid use in critically ill COVID-19 patients and the degree of lung involvement on chest CT. Furthermore, we aim to investigate the relationship of temporal changes in lung involvement both with the use of prolonged corticosteroid courses and with mortality. Another secondary aim of this study is to investigate whether chest CT reports influenced the duration of corticosteroid therapy. We present this article in accordance with the STROBE reporting checklist (available at https://jeccm.amegroups.com/article/view/10.21037/jeccm-2026-0011/rc).

Methods

Study design and participants

This study is a secondary and exploratory analysis of a previously published cohort including 1,539 adults with COVID-19 who were admitted to the Hospital de Clínicas de Porto Alegre, Brazil, from August 2020 to June 2021 (13). Inclusion criteria for the original cohort consisted of a positive polymerase chain reaction (PCR) test for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from nasal swabs and corticosteroid use during hospital stay. Patients with hospital stay shorter than 10 days, or individuals who died before 10 days of admission, were excluded. For the current analysis, only patients admitted to the intensive care unit (ICU) were considered. Patients without a CT scan and those whose first chest CT scan was performed more than 10 days after admission were also excluded. After applying these exclusion criteria, 218 patients remained eligible. The inclusion of all eligible patients was not feasible due to the resource-intensive nature of manual chart review and detailed chest CT assessment. In order to mitigate the risk of selection bias, eligible patients were randomly selected for the final analysis using the sample {base} function in R.

For the main analysis, subjects were separated into two groups according to the duration of corticosteroid therapy. Patients who received corticosteroids for up to 10 days were included in the short-course group, whereas patients who received corticosteroids for more than 10 days were included in the prolonged-course group. The cutoff of 10 days was defined according to the duration used in the main study that endorsed the use of corticosteroids for COVID-19 (11). For a secondary analysis, patients were divided into survivors and non-survivors.

Procedures

COVID-19 diagnosis was confirmed after a polymerase chain reaction test for SARS-CoV-2 using nasal swabs. After eligible patients were identified, electronic medical records were manually reviewed. Variables included demographic characteristics and laboratory data. Type, dosing, and duration of corticosteroid therapy were evaluated for all patients, and methylprednisolone-equivalent dosing was used for comparison. Accordingly, 1 mg of dexamethasone was considered equivalent to 5.33 mg of methylprednisolone, and 1 mg of hydrocortisone was considered equivalent to 0.2 mg of methylprednisolone. For each patient, the highest single dose administered during hospitalization was recorded and expressed as the corresponding methylprednisolone-equivalent dose (18). Data related to clinical outcomes [ICU and in-hospital mortality, duration of ICU and hospital stay, duration of mechanical ventilation, use of extracorporeal membrane oxygenation (ECMO), and need for prone positioning] were also obtained. The Simplified Acute Physiology Score 3 (SAPS 3) was used to assess disease severity (19).

CT examinations were performed using a 64-channel multidetector scanner (Aquilion 64, Toshiba Medical Systems, Japan or Revolution, GE Medical Systems, Chicago, Illinois, USA), in the supine position, during maximum inspiration. All acquisitions were volumetric (slice thickness of 1–2 mm) and reconstructed using a high spatial frequency algorithm. Images were stored and analyzed with an Image Communication and Archiving System (Enterprise Imaging, Agfa HealthCare, Mortsel, Belgium). Imaging analysis was conducted independently by two radiologists, using a visual semi-quantitative method that scores 0 to 5 points according to the percentage of involvement in each lung lobe for the three elementary patterns associated with COVID-19: ground-glass opacities, consolidations, and crazy-paving. The score for each elementary lesion ranged from 0 to 25. The total CT score was calculated summing individual scores for each lobe and elementary pattern, ranging from a minimum of 0 to a maximum of 28 points, with higher scores indicating more severe lung involvement (20). For patients who underwent a second chest CT during hospitalization, the variation in CT scores between the first and second chest CT (delta CT score) was computed to assess the degree of lung involvement over time. The delta CT score is defined as the total CT score for the second chest CT minus the total CT score for the first chest CT.

Additionally, the senior radiologist manually reviewed all image reports to identify terms indicating either improvement, stability, or worsening of patterns (see Table S1).

Outcomes

The primary outcome was the degree of lung involvement on chest CT measured by the visual semi-quantitative scoring system described above. Secondary outcomes included the difference in total chest CT score between the first and the second chest CT (delta CT score), as well as the frequency of key imaging reports indicating worsening, stability, or improvement of lung involvement over time.

Statistical analysis

Continuous variables are presented as mean ± standard deviation (SD) or median [interquartile range (IQR)], as appropriate. Categorical variables are presented as counts and percentages. Baseline characteristics were summarized using the TableOne package (version 0.13.2; https://github.com/kaz-yos/tableone). Group comparisons for primary and secondary outcomes were performed using one-way analysis of variance (oneway.test) for normally distributed continuous variables and the Kruskal-Wallis test (kruskal.test) for non-normally distributed continuous variables; for comparisons involving two groups, these tests are equivalent to the Student’s t-test and Mann-Whitney U test, respectively. Normality was assessed by visual inspection of distribution plots and the Kolmogorov-Smirnov test. Categorical variables were compared using the Chi-squared test. Three separate multivariable regression models were constructed for each outcome. In the first model, the outcome was in-hospital mortality, the exposure of interest was the delta CT score, and it was adjusted for admission C-reactive protein and SAPS 3. In the second model, the outcome was also in-hospital mortality, the exposure of interest was the delta CT score, and it was adjusted for maximal methylprednisolone dose and SAPS 3. In the third model, the outcome was prolonged corticosteroid use, and the exposure of interest was the ground-glass opacity score from the first chest CT, and it was adjusted for SAPS 3 and admission C-reactive protein. A Pearson’s correlation coefficient was calculated to assess the linear relationship between CT scores for both radiologists. All CT scans were independently evaluated by two radiologists using the predefined CT severity scoring system. Inter-reader reliability was assessed using a two-way mixed-effects intraclass correlation coefficient for absolute agreement and single measurements {intraclass correlation coefficient (ICC)[A, 1]}. Agreement between readers was further explored using Bland-Altman analysis. Scores obtained from the senior radiologist were used for the final analysis. Statistical analyses were performed using the R software version 4.2.2 (R Foundation for Scientific Computing). For all analyses, differences were statistically significant at P values <0.05.

Ethical consideration

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of Hospital de Clínicas de Porto Alegre (project No. 2021-0220), and the study was registered on the Plataforma Brasil (project No. 46985921.0.1001.5327). The Institutional Ethics Committee of Hospital de Clínicas de Porto Alegre waived the need to obtain informed consent for the collection, analysis, and publication of the retrospectively obtained and anonymized data for this non-interventional study.

Results

A total of 116 patients admitted to the ICU fulfilled the inclusion criteria (Figure 1). Of these, six patients were excluded: four were admitted for reasons unrelated to COVID-19 and subsequently developed a COVID-19 infection during hospitalization, whereas two tested positive for COVID-19 on screening but remained asymptomatic. The final analysis included 110 critically ill patients, 63% male, with a mean age of 54±15 years and a mean SAPS 3 score of 59±14, which was similar between groups. From those, 54 (49%) received a short course of corticosteroids, and 56 (51%) a prolonged course. Patients in the prolonged-course group received corticosteroids for a median of 22 (IQR, 14–33) days, compared to 9 (IQR, 7–10) days for patients in the short-course group (P<0.001). The prolonged-course group received higher maximum methylprednisolone-equivalent doses of corticosteroids compared to the short-course group [59 (IQR, 40–120) vs. 32 (IQR, 32–32) mg, P<0.001]. Table 1 summarizes the baseline characteristics of the studied population.

Figure 1 Flowchart of patient enrolment and reasons for exclusion. COVID-19, coronavirus disease 2019; CT, computed tomography; ICU, intensive care unit.

Patients underwent the first chest CT scan after a median of 3 days of hospital stay and after a median of 2 days of corticosteroid therapy. Three patients underwent the first chest CT before initiation of corticosteroid therapy. A total of 75 patients (68%) underwent a second chest CT scan during hospitalization, after a median of 14 days. As expected, patients who underwent at least two chest CT scans had longer hospital LOS, longer duration of mechanical ventilation, and received prolonged corticosteroid courses more frequently (Table S2).

Degree of lung involvement on computed chest CT

There was a strong correlation between lung involvement scores assessed by both radiologists (r=0.80, P<0.001). Interobserver agreement between the two radiologists was evaluated using Bland-Altman analysis. The mean difference (bias) was −2.85, with 95% limits of agreement from −11.87 to 6.17 (Figure S1). The intraclass correlation coefficient for absolute agreement and single measurements {ICC[A, 1]} was 0.73 [95% confidence interval (CI): 0.47−0.85].

Table 2 summarizes the degree of lung involvement on chest CT between groups categorized according to the duration of corticosteroid therapy. At the first CT scan, patients had a median total CT score of 20 (IQR, 14–25). Even though there was a trend towards higher total CT scores in the prolonged-course group, this difference was not statistically significant [median 21 (IQR, 16–27) vs. 17 (IQR, 13–24), P=0.054]. The prolonged-course group exhibited higher scores for ground-glass opacities compared to the short-course group [15 (IQR, 9–19) vs. 10 (IQR, 4–15), P=0.004]. There was no difference in CT scores for consolidation and crazy-paving between groups.

In order to address possible baseline imbalances contributing to the differences in the degree of lung involvement between the short- and prolonged-groups, we constructed logistic regression models adjusting for markers of illness severity and baseline characteristics. After adjustment, the total CT score for ground-glass opacities on the first CT scan [odds ratio (OR) =1.11; 95% CI: 1.04–1.18; P=0.001] and C-reactive protein (OR =0.99; 95% CI: 0.988–0.999; P=0.03) were significantly associated with prolonged corticosteroid use, whereas SAPS 3 was not significantly associated (OR =1.00; 95% CI: 0.97–1.04; P=0.80). At the second CT scan, the median total CT score was 19 (IQR, 10–25), with a higher degree of lung involvement among patients who received prolonged courses of corticosteroids compared to short courses [20 (IQR, 13–25) vs. 13 (IQR, 8–21), P=0.02]. Scores for the three elementary patterns did not differ significantly between groups. There was a median decrease of two points (IQR, 6 to 1) in the total CT score between the first and the second chest CT (delta CT score). The delta CT score did not differ between the prolonged and short-course groups.

Table 3 compares chest CT scores between survivors and non-survivors. There was no significant difference in the total CT score or in any elementary pattern between the first and the second chest CT scans when comparing survivors and non-survivors. The delta CT score did not differ between survivors and non-survivors [−3 (IQR, −7 to 1) vs. 0 (IQR −3 to 1), P=0.07]. However, we performed a multiple logistic regression to investigate the association between in-hospital mortality and the delta CT score. For every 1-point worsening in the total CT score, there was a trend of increase in the odds of death by 9%, which did not reach statistical significance (OR =1.09; 95% CI: 1.00–1.20; P=0.07). The other variables included in the model were SAPS 3, for which a positive association was found (OR =1.04; 95% CI: 1.04–1.09; P=0.041), and C-reactive protein levels (OR =1.01; 95% CI: 1.00–1.01; P=0.12). This trend of association between in-hospital mortality and delta CT score was also observed in a second model (OR =1.09; 95% CI: 1.00–1.20; P=0.06), after adjusting for SAPS 3 (OR =1.04; 95% CI: 1.00–1.09; P=0.06) and maximal methylprednisolone equivalent dosing (OR =1.00; 95% CI: 0.99–1.00; P=0.04).

A review of the imaging reports revealed no significant differences between the short- and prolonged-course groups in the occurrence of key terms regarding the evolution of lung changes. Figure 2 displays a lollipop chart of the terms used in radiological reports for the prolonged and short-course groups.

Figure 2 Lollipop chart comparing the frequency of terms used in radiology reports between patients in the short vs. prolonged-course groups.

Exploratory outcomes

Table S3 summarizes the clinical outcomes. The overall in-hospital and ICU mortality were 26.4% and 23.6%, respectively, and were similar between groups. Time spent on mechanical ventilation, as well as length of hospital and ICU stays, were longer for patients in the prolonged-course group. The need for prone positioning and ECMO was not associated with the duration of corticosteroid courses.

Discussion

This retrospective longitudinal study investigated the association between the degree of lung involvement in chest CT and duration of corticosteroid courses in critically ill COVID-19 patients. We reported that more severe degrees of lung involvement at different time points are associated with prolonged use of corticosteroids. Moreover, we found a trend of association between temporal changes in lung involvement and mortality, with an increase in the risk of death associated with worsening in lung involvement over time.

To our knowledge, the present study is the first to address the relationship between corticosteroid therapy duration and lung involvement in COVID-19. Prior studies have demonstrated that a greater extent of lung opacities is associated with higher inflammatory burden and worse clinical outcomes (3,4). In this context, the greater lung involvement observed in our study may help explain the longer duration of corticosteroid therapy, possibly reflecting attempts to control a persistent inflammatory response. This interpretation is supported by evidence showing improvement in lung involvement among COVID-19 patients treated with corticosteroids compared with those who did not receive this therapy (21), as well as by the evidence that improvement in lung lesions is associated with positive clinical outcomes, such as successful weaning from mechanical ventilation (22). However, the retrospective design precludes causal inference, and the observed association may reflect confounding by indication, whereby patients with more severe disease and greater CT involvement were more likely to receive prolonged corticosteroid therapy. Although the delta total CT score showed a trend toward prognostic relevance for hospital mortality after adjustment for SAPS 3, and admission C-reactive protein, these analyses do not resolve causal ambiguity regarding treatment duration. A reduction in lung involvement over time, as reflected by changes in the delta CT score, was observed, with no clear differences according to corticosteroid therapy duration. Not all eligible patients underwent a second chest CT during hospitalization, introducing potential attrition bias.

In a previous study, we demonstrated an increase in mortality among patients with COVID-19 pneumonia who received prolonged corticosteroid courses (13). In the present study, we were not able to replicate this finding—possibly due to the smaller sample size and the lack of power to detect such difference. However, we demonstrated a trend of association between the evolution of lung involvement over time and mortality, with survivors presenting a more pronounced improvement in chest CT than non-survivors. Even though these findings did not reach statistical significance, they suggest that lung repair may have a critical role in the prognosis of critically ill COVID-19 patients.

Since clinical decisions to prolong corticosteroid therapy may have been more influenced by the final radiological report than the imaging itself, we analyzed the key terms included in these reports. However, our analysis revealed no significant differences in key terms between the short and prolonged-course groups. The lack of significant difference between radiological reports in both groups suggests that the decision to prolong corticosteroid therapy was not based solely on radiological reports and findings, but rather on multiple measured and unmeasured factors, including disease severity and clinical course, laboratory results, and the potential risks and benefits of continuing corticosteroid treatment.

In this study, we used a radiologist-based, lobar-based semi-quantitative lung assessment tool to determine the radiological image burden of COVID-19 pneumonia. This scoring system was as accurate as other semi-quantitative or quantitative methods to assess the volume of lung opacities (23). Semi-quantitative methods have been demonstrated to have great interobserver agreement and predict clinical severity (23,24). However, such methods are limited by subjectivity and require significant interpretation time. On the other hand, artificial intelligence-based methods offer greater accuracy, efficiency, and reliability in the evaluation of lung involvement among patients with COVID-19 pneumonia (25,26). This is especially true for mild to moderate cases, which were not the focus of the present study. And despite their growing relevance in the evaluation of chest imaging, there is no evidence of benefit in patient-centered outcomes with the use of artificial intelligence-based methods in comparison to semi-quantitative methods for patients with COVID-19 (27,28).

There are certain limitations to consider. As this study was performed in one center, generalizability may be impaired. Generalizability is also impaired by the lack of specific criteria to define severe COVID-19. Due to the limited sample size, the study may have been underpowered to detect differences in outcomes, particularly in mortality. Furthermore, an important limitation of this study is that not all eligible patients were included in the analysis, owing to the substantial effort required to quantify elementary lesions on CT scans. Consequently, the possibility of selection bias cannot be excluded. However, the random selection from a previously published dataset might mitigate this risk and ensure generalizability of the results. Even though chest CT analysis was conducted by two radiologists, we used the senior scores for analysis. This introduces subjectivity and may reduce the robustness of the results. Moreover, the review of radiology reports was conducted by only one of the radiologists, which was unaware of group assignments. This study was observational and exploratory in nature and was designed to generate hypotheses rather than establish causal relationships. Reverse causation cannot be excluded, and it is possible that greater overall clinical severity may have influenced corticosteroid therapy rather than imaging findings alone. Besides, the clinical significance of temporal CT findings is questionable in light of the patient’s clinical evolution. Nevertheless, it is noteworthy that the association between ground-glass opacity on the first chest CT and prolonged corticosteroid use was maintained even under more conservative analytical approaches and after adjustment for key clinical variables. Furthermore, due to the observational nature of the study, not all eligible patients underwent a second chest CT during hospitalization, which creates a risk of attrition bias for analyses of the temporal evolution of imaging findings, and the interval between first and second chest CT was not standardized, which might influence the magnitude of radiological changes observed.

Conclusions

In conclusion, although there is limited evidence to support corticosteroid therapy beyond 10 to 14 days in critically ill patients with COVID-19, patients with more extensive lung involvement appeared more likely to receive prolonged corticosteroid courses. Moreover, temporal changes in lung involvement on chest CT showed a trend toward an association with mortality. The findings of our study reinforce the prognostic value of the chest CT in patients with COVID-19 pneumonia. Further prospective studies are needed to confirm these findings and clarify causality. Besides, future advances in artificial intelligence-based approaches may enhance the precision and reproducibility of quantifying lung involvement in patients with COVID-19.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jeccm.amegroups.com/article/view/10.21037/jeccm-2026-0011/rc

Data Sharing Statement: Available at https://jeccm.amegroups.com/article/view/10.21037/jeccm-2026-0011/dss

Peer Review File: Available at https://jeccm.amegroups.com/article/view/10.21037/jeccm-2026-0011/prf

Funding: This work was supported by Programa de Financiamento e Incentivo à Pesquisa do Hospital de Clínicas de Porto Alegre (FIPE-HCPA, project No. 2021-0220).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jeccm.amegroups.com/article/view/10.21037/jeccm-2026-0011/coif). C.T. serves as an unpaid editorial board member of Journal of Emergency and Critical Care Medicine from December 2022 to December 2027. The other 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 the Institutional Ethics Committee of Hospital de Clínicas de Porto Alegre (project No. 2021-0220), and the study was registered on the Plataforma Brasil (project No. 46985921.0.1001.5327). The Institutional Ethics Committee of Hospital de Clínicas de Porto Alegre waived the need to obtain informed consent for the collection, analysis, and publication of the retrospectively obtained and anonymized data for this non-interventional study.

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. Machnicki S, Patel D, Singh A, et al. The Usefulness of Chest CT Imaging in Patients With Suspected or Diagnosed COVID-19: A Review of Literature. Chest 2021;160:652-70. [Crossref] [PubMed]
2. Kwee TC, Kwee RM. Chest CT in COVID-19: What the Radiologist Needs to Know. Radiographics 2020;40:1848-65. [Crossref] [PubMed]
3. Krisht AAH, Grapin K, de Beauchene RC, et al. SARS-CoV2 pneumonia patients admitted to the ICU: Analysis according to clinical and biological parameters and the extent of lung parenchymal lesions on chest CT scan, a monocentric observational study. PLoS One 2024;19:e0308014. [Crossref] [PubMed]
4. Torres-Ramirez CA, Timaran-Montenegro D, Mateo-Camacho YS, et al. CT-based pathological lung opacities volume as a predictor of critical illness and inflammatory response severity in patients with COVID-19. Heliyon 2022;8:e11908. [Crossref] [PubMed]
5. Wu C, Chen X, Cai Y, et al. Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease 2019 Pneumonia in Wuhan, China. JAMA Intern Med 2020;180:934-43. [Crossref] [PubMed]
6. Sturgill JL, Mayer KP, Kalema AG, et al. Post-intensive care syndrome and pulmonary fibrosis in patients surviving ARDS-pneumonia of COVID-19 and non-COVID-19 etiologies. Sci Rep 2023;13:6554. [Crossref] [PubMed]
7. de Souza Xavier Costa N, Ribeiro Júnior G, do Nascimento ECT, et al. COVID-19 induces more pronounced extracellular matrix deposition than other causes of ARDS. Respir Res 2023;24:281. [Crossref] [PubMed]
8. Razzaque MS, Taguchi T. Pulmonary fibrosis: cellular and molecular events. Pathol Int 2003;53:133-45. [Crossref] [PubMed]
9. Hama Amin BJ, Kakamad FH, Ahmed GS, et al. Post COVID-19 pulmonary fibrosis; a meta-analysis study. Ann Med Surg (Lond) 2022;77:103590. [Crossref] [PubMed]
10. Siemieniuk RA, Bartoszko JJ, Zeraatkar D, et al. Drug treatments for covid-19: living systematic review and network meta-analysis. BMJ 2020;370:m2980. [Crossref] [PubMed]
11. Horby P, Lim WS, Emberson JR, et al. Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med 2021;384:693-704.
12. Tomazini BM, Maia IS, Cavalcanti AB, et al. Effect of Dexamethasone on Days Alive and Ventilator-Free in Patients With Moderate or Severe Acute Respiratory Distress Syndrome and COVID-19: The CoDEX Randomized Clinical Trial. JAMA 2020;324:1307-16. [Crossref] [PubMed]
13. Viana MV, Pellegrini JAS, Perez AV, et al. Association between prolonged corticosteroids use in COVID-19 and increased mortality in hospitalized patients: a retrospective study with inverse probability of treatment weighting analysis. Crit Care 2023;27:143. [Crossref] [PubMed]
14. Rhen T, Cidlowski JA. Antiinflammatory action of glucocorticoids--new mechanisms for old drugs. N Engl J Med 2005;353:1711-23. [Crossref] [PubMed]
15. Landolf KM, Lemieux SM, Rose C, et al. Corticosteroid use in ARDS and its application to evolving therapeutics for coronavirus disease 2019 (COVID-19): A systematic review. Pharmacotherapy 2022;42:71-90. [Crossref] [PubMed]
16. Ljilja Posavec A, Cvetković Kučić D, Zagorec N, et al. Prolonged corticosteroid therapy and lung abnormalities in patients after severe COVID-19 pneumonia. Sarcoidosis Vasc Diffuse Lung Dis 2024;41:e2024052. [Crossref] [PubMed]
17. Roël M, Schandl A, Jonmarker S, et al. Corticosteroids and long-term pulmonary function after critical illness due to COVID-19- a single-center cohort study. BMC Pulm Med 2025;25:201. [Crossref] [PubMed]
18. Asare K. Diagnosis and treatment of adrenal insufficiency in the critically ill patient. Pharmacotherapy 2007;27:1512-28. [Crossref] [PubMed]
19. Moreno RP, Metnitz PG, Almeida E, et al. SAPS 3--From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med 2005;31:1345-55.
20. Pan F, Ye T, Sun P, et al. Time Course of Lung Changes at Chest CT during Recovery from Coronavirus Disease 2019 (COVID-19). Radiology 2020;295:715-21. [Crossref] [PubMed]
21. Calabrese C, Pafundi PC, Mollica M, et al. Effectiveness of corticosteroids on chest high-resolution computed tomography features of COVID-19 pneumonia. Ther Adv Respir Dis 2021;15:17534666211042533. [Crossref] [PubMed]
22. Cigolini MP, Schneider B, Bellé NL, et al. Association between chest tomographic changes and weaning from mechanical ventilation in patients with severe acute respiratory syndrome due to SARS-CoV-2: a retrospective cohort study. J Emerg Crit Care Med 2026;10:1.
23. Guan X, Yao L, Tan Y, et al. Quantitative and semi-quantitative CT assessments of lung lesion burden in COVID-19 pneumonia. Sci Rep 2021;11:5148. [Crossref] [PubMed]
24. Inoue A, Takahashi H, Ibe T, et al. Comparison of semiquantitative chest CT scoring systems to estimate severity in coronavirus disease 2019 (COVID-19) pneumonia. Eur Radiol 2022;32:3513-24. [Crossref] [PubMed]
25. Szabó M, Kardos Z, Kostyál L, et al. The importance of chest CT severity score and lung CT patterns in risk assessment in COVID-19-associated pneumonia: a comparative study. Front Med (Lausanne) 2023;10:1125530. [Crossref] [PubMed]
26. Atceken Z, Celik Y, Atasoy C, et al. The Diagnostic Utility of Artificial Intelligence-Guided Computed Tomography-Based Severity Scores for Predicting Short-Term Clinical Outcomes in Adults with COVID-19 Pneumonia. J Clin Med 2023;12:7039. [Crossref] [PubMed]
27. Grodecki K, Killekar A, Simon J, et al. Artificial intelligence-assisted quantification of COVID-19 pneumonia burden from computed tomography improves prediction of adverse outcomes over visual scoring systems. Br J Radiol 2023;96:20220180. [Crossref] [PubMed]
28. Baldi D, Punzo B, Colacino A, et al. Severity assessment of COVID-19 disease: radiological visual score versus automated quantitative CT parameters using a pneumonia analysis algorithm. Front Med (Lausanne) 2025;12:1606771. [Crossref] [PubMed]