Despite significant strides in big data technology, extracting information from unstructured clinical data remains a formidable challenge. This study investigated the utility of large language models (LLMs) for extracting clinical data from unstructured radiological reports without additional training. In this retrospective study, 1800 radiologic reports, 600 from each of the three university hospitals, were collected, with seven pulmonary outcomes defined. Three pulmonology-trained specialists discerned the presence or absence of diseases. Data extraction from the reports was executed using Google Gemini Pro 1.0, OpenAI’s GPT-3.5, and GPT-4. The gold standard was predicated on agreement between at least two pulmonologists. This study evaluated the performance of the three LLMs in diagnosing seven pulmonary diseases (active tuberculosis, emphysema, interstitial lung disease, lung cancer, pleural effusion, pneumonia, and pulmonary edema) utilizing chest radiography and computed tomography scans. All models exhibited high accuracy (0.85–1.00) for most conditions. GPT-4 consistently outperformed its counterparts, demonstrating a sensitivity of 0.71–1.00; specificity of 0.89–1.00; and accuracy of 0.89 and 0.99 across both modalities, thus underscoring its superior capability in interpreting radiological reports. Notably, the accuracy of pleural effusion and emphysema on chest radiographs and pulmonary edema on chest computed tomography scans reached 0.99. The proficiency of LLMs, particularly GPT-4, in accurately classifying unstructured radiological data hints at their potential as alternatives to the traditional manual chart reviews conducted by clinicians.

Backgrounds The management of lung complications, especially fibrosis, after COVID-19 pneumonia, is an important issue in the COVID-19 post-pandemic era. We aimed to investigate risk factors for pulmonary fibrosis development in patients with severe COVID-19 pneumonia. Methods Clinical and radiological data were prospectively collected from 64 patients who required mechanical ventilation due to COVID-19 pneumonia and were enrolled from eight hospitals in South Korea. Fibrotic changes on chest CT were evaluated by visual assessment, and extent of fibrosis (mixed disease score) was measured using automatic quantification system. Results 64 patients were enrolled, and their mean age was 58.2 years (64.1% were males). On chest CT (median interval: 60 days [IQR; 41–78 days] from enrolment), 35 (54.7%) patients showed ≥3 fibrotic lesions. The most frequent fibrotic change was traction bronchiectasis (47 patients, 73.4 %). Median extent of fibrosis measured by automatic quantification was 10.6% (IQR, 3.8–40.7%). In a multivariable Cox proportional hazard model, which included nine variables with a p value of <0.10 in an unadjusted analysis as well as age, sex and Body Mass Index, male sex (HR, 3.01; 95% CI, 1.27 to 7.11) and higher initial Sequential Organ Failure Assessment (SOFA) score (HR, 1.18; 95% CI, 1.02 to 1.37) were independently associated with pulmonary fibrosis (≥3 fibrotic lesions). Conclusion Our data suggests that male gender and higher SOFA score at intensive care unit admission were associated with pulmonary fibrosis in patients with severe COVID-19 pneumonia requiring mechanical ventilation.

Background: Incremental shuttle walking tests (ISWT) are regarded as valuable alternatives to 6-min walking tests (6MWT) and cardiopulmonary exercise tests (CPET) owing to the maximal and externally paced loading. This study investigated the validity and reliability of ISWT by analyzing the correlation of the distances of two field tests with peak oxygen consumption (VO₂) of CPET in patients with COPD. Methods: In this randomized controlled trial, patients with COPD were enrolled from two hospitals. Three assessments were performed for all patients. The ISWT and 6MWT were repeated twice in Hospital 1 to assess reliability. Results: A total of 29 patients were enrolled. The distances of ISWT (0.782, p < 0.001) and 6MWT (0.512, p = 0.005) correlated with peak VO₂. The intraclass correlation coefficients of both ISWT (0.988, p < 0.001) and 6MWT (0.959, p < 0.001) was high. Patients with higher peak VO₂ walked a longer distance in ISWT than 6MWT (r = 0.590, p < 0.001). Conclusions: The ISWT more highly correlates with peak VO₂ than the 6MWT and has excellent reliability in patients with COPD. According to peak VO₂, the walking distances of each field test varied, suggesting that the application should be personalized for the exercise capacity.

Introduction: Coronavirus disease 2019 (COVID-19) still has a high mortality rate when it is severe. Regdanvimab (CT-P59), a neutralizing monoclonal antibody that has been proven effective against mild to moderate COVID-19, may be effective against severe COVID-19. This study was conducted to determine the effectiveness of the combined use of remdesivir and regdanvimab in patients with severe COVID-19. Methods: From March to early May 2021, 124 patients with severe COVID-19 were admitted to Ulsan University Hospital (Ulsan, Korea) and received oxygen therapy and remdesivir. Among them, 25 were also administered regdanvimab before remdesivir. We retrospectively compared the clinical outcomes between the remdesivir alone group [n = 99 (79.8%)] and the regdanvimab/remdesivir group [n = 25 (20.2%)]. Results: The oxygen-free days on day 28 (primary outcome) were significantly higher in the regdanvimab/remdesivir group [mean ± SD: 19.36 ± 7.87 vs. 22.72 ± 3.66, p = 0.003]. The oxygen-free days was also independently associated with use of regdanvimab in the multivariate analysis, after adjusting for initial pulse oximetric saturation (SpO₂)/fraction of inspired oxygen (FiO₂) ratio (severity index). Further, in the regdanvimab/remdesivir group, the lowest SpO₂/FiO₂ ratio during treatment was significantly higher (mean ± SD: 237.05 ± 89.68 vs. 295.63 ± 72.74, p = 0.003), and the Kaplan-Meier estimates of oxygen supplementation days in surviving patients (on day 28) were significantly shorter [mean ± SD: 8.24 ± 7.43 vs. 5.28 ± 3.66, p = 0.024]. Conclusions: In patients with severe COVID-19, clinical outcomes can be improved by administering regdanvimab, in addition to remdesivir.

Kwon OB, et al. J Korean Med Sci. 2025 Jan;40:e73. https://doi.org/10.3346/jkms.2025.40.e73

Background: Screening patients with asymptomatic active tuberculosis (TB) is crucial as they can transmit the disease. Identifying the risk factors for transmission is essential for targeted screening. Understanding how the infectiousness of asymptomatic patients with TB affects disease outcomes is crucial for developing strategies to control TB spread. Methods: We analyzed the national Korean TB cohort data to determine the factors associated with transmission risk and clinical outcomes in patients with asymptomatic pulmonary TB. The primary outcome was the factors associated with a risk factor for transmission, while the secondary outcome was mortality in asymptomatic patients with pulmonary TB stratified by transmission risk. Results: Among 20 455 patients with pulmonary TB, 7434 (36.4%) were asymptomatic, while 1520 (25.5%) had potential transmission risks, indicated by a positive sputum acid-fast bacillus smear test or cavitation on chest radiographs. The factors associated with a higher transmission risk included male sex (odds ratio [OR], 1.385; 95% CI, 1.172–1.636; P < .001), low body mass index (BMI; OR, 1.687; 95% CI, 1.420–2.004; P < .001), current smoking (OR, 1.443; 95% CI, 1.213–1.716; P < .001), diabetes (OR, 1.399; 95% CI, 1.201–1.629; P < .001), and autoimmune disease (OR, 2.233; 95% CI, 1.295–3.850; P = .004). The mortality rate was higher in patients with a risk factor for transmission risk than in those without (9.3 vs 7.1%; P = .008). Conclusions: Lean, smoking men with asymptomatic TB who have DM and/or autoimmune diseases have higher transmission and mortality risk. Asymptomatic populations with these risk factors warrant targeted screening.