Ventilator-associated pneumonia (VAP) is driven in part by microaspiration along the endotracheal tube cuff, a process difficult to measure directly in ventilated patients. Because microbial communities differ between airway compartments, changes in their similarity over time may serve as an indirect readout of microaspiration. We assessed whether the Venner PneuX^® Tube (VT) system, combining cuff-pressure monitoring, subglottic suction, and a biofilm-resistant coating, reduces microbial exchange between airway compartments compared with a standard endotracheal tube (ST).

In a prospective, randomized, single-center pilot study, 50 adults with acute respiratory failure received an ST or VT intubation. Microbial communities from five airway niches (throat, tracheal secretions, right upper, right lower, and left lower lung lobes) were sampled by 16 S rRNA sequencing at intubation (T1), after four days of ventilation (T2) and, for the tube tip, at extubation (T3). The primary outcome was the change in beta diversity (Morisita-Horn distances) between tube-associated, upper-airway, and lower-airway communities from T1 to T2.

Twenty-one of 50 randomized patients had complete microbiota data sets (9 ST, 12 VT) and were comparable in demographics, comorbidities, severity, and ventilation duration. In the ST group, tube-associated communities became more similar to tracheal and lower-airway communities from T1 to T2 (e.g. TS-Tube Morisita-Horn 0.55 → 0.30, p = 0.001), while lung regions diverged from each other and from the throat (LLL-LRL 0.08 → 0.31, p = 0.003; LLL-Throat 0.30 → 0.61, p < 0.001). None of these distances changed significantly in the VT group. Lower-airway Shannon diversity declined in both groups, more in the VT group.

Standard intubation produced progressive microbial convergence between airway compartments, while the VT system did not. The findings provide biological plausibility for previously reported VAP reductions with the VT system and show that microbiota sampling can detect device-related differences in airway community structure, warranting further investigation as an endpoint for evaluating airway devices.

The study is registered in the German Clinical Trials Register (DRKS- Deutsches Register für klinische Studien) under the clinical trial number: DRKS00029176. The Date of Trial Registration was 07.07.2022.

During the COVID-19 pandemic, inter-hospital transfer of critically ill patients between intensive care units was widely used to manage capacity shortages. While several individual studies have compared outcomes of transferred and non-transferred patients, no systematic synthesis of this evidence exists. This systematic review aims to determine whether inter-hospital ICU-to-ICU transfer of adult COVID-19 patients is associated with increased mortality or other adverse clinical outcomes. We systematically searched PubMed, Web of Science, and Scopus for observational studies comparing clinical outcomes of adult COVID-19 patients who underwent inter-hospital ICU-to-ICU transfer with non-transferred ICU patients. Two reviewers screened, selected, and extracted data independently and in duplicate. Risk of bias was assessed using the Newcastle-Ottawa Scale and ROBINS-I, and certainty of evidence using GRADE. Exploratory random-effects meta-analyses were performed separately for each effect measure, with Hartung-Knapp and crude-effect sensitivity analyses. Nine observational studies from seven countries were included (approximately 6,100 transferred and 29,200 non-transferred ICU patients). None of the adjusted effect estimates showed a statistically significant mortality disadvantage for transferred patients. Pooled adjusted subgroup estimates were an odds ratio of 1.29 (95% CI 0.84 to 1.98; k = 3) and a hazard ratio of 0.71 (95% CI 0.34 to 1.46; k = 2). Length of stay was longer in transferred patients in most studies. The available observational evidence did not show a clear increase in mortality among transferred compared with non-transferred COVID-19 ICU patients. These findings support the use of inter-hospital transfer as a safe strategy for managing ICU surge capacity. PROSPERO CRD420261356915.

To externally validate the RSNA 2020 challenge 2^nd-place deep-learning (DL) algorithm for detecting pulmonary embolism (PE) on computed tomography pulmonary angiography (CTPA) scans with additional focus on subsegmental-only PEs (SSPE). Between 2015 and 2022 1,038 CTPAs were retrospectively enrolled. An experienced radiologist (> five years) labelled CTPAs for the presence, location (central, lobar and segmental, subsegmental-only) and side (right-sided, left-sided, bilateral) of PE. The DL model was tested for its ability to predict these labels by analyzing accuracy, sensitivity, specificity and area under the receiver operating characteristic curve (AUROC) using different model outcome probabilities for each analysis. Overall, 136 central, 375 peripheral PEs (303 lobar and segmental PEs, 72 SSPEs), and 527 patients without any PE were analyzed. The model correctly predicted the presence of any PE in 921/1,038 patients (88.7%), yielding a sensitivity of 0.80, specificity of 0.97 and AUROC of 0.94. No central PE was missed, whereas 100/375 (26.7%) peripheral PEs remained undetected. Using the corresponding model output probability, the central PE status was correctly identified in 999/1,038 (96.2%) patients, while peripheral PE was correctly identified in 785/1,038 (75.6%) patients. This corresponded to sensitivities of 0.94 and 0.77, specificities of 0.97 and 0.75, and AUROC values of 0.99 and 0.74, respectively. The model performed better in detecting right-sided compared to left-sided PEs (AUROC: right- vs. left-sided: 0.95 and 0.92, p < 0.05). SSPE status was correctly predicted in 796/1,038 (76.7%) patients, yielding a sensitivity of 0.94, a specificity of 0.22, and an AUROC of 0.48, respectively. The model demonstrated high performance in detecting any PE, performing best on central and slightly lower on lobar and segmental PE. Predicting the presence of SSPEs was difficult, likely due to the model not being explicitly trained for this subtype.

Bronchiolitis obliterans syndrome (BOS), a severe complication following allogeneic hematopoietic stem cell transplantation (allo-HSCT), presents as chronic graft-versus-host disease characterized by inflammation and fibrosis of the small airway epithelium. This progressive fibrosis obstructs bronchiolar airways, leading to respiratory distress in patients. Due to the lack of effective treatments, a deeper understanding of the underlying mechanisms is crucial. This study explored the molecular mechanisms underlying the abnormal glycosylation regulation in BOS, aiming to provide new insights for early diagnosis and treatment of this disease.

Bronchoalveolar lavage fluid (BALF) was collected from patients with and without BOS following allo-HSCT. N-glycans from the BALF were enriched using a solid-phase extraction method. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF/TOF-MS) was used to detect N-glycosylation characteristics in the BALF of BOS patients. This phenomenon was validated using a graft-versus-host disease (GVHD) mouse model. The impact of MGAT3 knockdown on the biological functions of airway epithelial cells (BEAS-2B) was then examined using lentivirus transfection. The molecular mechanism by which the loss of MGAT3 enhances TGF-β signaling was explored using western blotting, enzyme-linked immunosorbent assay (ELISA), and immunofluorescence (IF).

Our study investigated N-glycosylation changes in BALF of BOS patients and identified a strong correlation between the loss of bisecting-GlcNAc N-glycans and BOS progression. We found a novel mechanism where MGAT3, the enzyme synthesizing bisecting-GlcNAc structures, critically regulates TGF-β signaling. MGAT3 and bisecting-GlcNAc deficiency significantly enhance TGF-β signaling by increasing TGF-β storage through upregu-lation of latent TGF-β binding protein 1 (LTBP1) and by increasing the availability of TGF-β receptors.

Our discovery highlights the crucial role of aberrant glycosylation in BOS progression and holds significant promise for developing novel biomarkers for early diagnosis, identifying new therapeutic targets, and ultimately improving the quality of life for BOS patients.

Diverse quarantine measures against the coronavirus disease 2019 (COVID-19) pandemic have affected the distribution patterns of respiratory pathogens. This study analyzed changes in the incidence and distribution of major respiratory pathogens during the COVID-19 pandemic and after the endemic declaration. In addition, the study aimed to provide a basis for future community-centered infectious disease response strategies by identifying co-infection patterns and whether pathogens re-emerge after mitigation.

In this retrospective study, the results of 105,619 respiratory pathogen tests, including FilmArray® Respiratory Panel (BioFire Diagnostics) assay (FilmArray RP), influenza tests, and COVID-19 tests, were collected at a Korean university hospital during the COVID-19 pandemic (Period I: March 2021 through April 2023) and endemic (Period II: May 2023 through February 2025). The changes in the distribution of major pathogens, age and gender characteristics, seasonality, and co-infection patterns were compared between the two periods.

The overall positivity rate for the FilmArray RP significantly increased from 64.4% in Period I to 71.1% in Period II (p < 0.001). In the age-specific analysis, age group 3 (4 months to 2 years) had the highest positivity rate in Period II (82.7%). Human rhinovirus/enterovirus was the most frequently detected pathogen, whereas respiratory syncytial virus cases increased in number, but decreased in proportion to all positive cases in Period II. The prevalence of adenoviruses and Mycoplasma pneumoniae has significantly increased since the pandemic. Human metapneumovirus has shifted its seasonality from autumn to summer and spring, whereas parainfluenza virus 3 has shifted from autumn to spring and summer. Among all positive cases, co-infections significantly increased to 17.6% and 22.3% during Periods I and II, respectively (p < 0.001).

This study revealed significant changes in the incidence, seasonality, age characteristics, and co-infection patterns of respiratory pathogens after the mitigation of COVID-19. These changes result from the re-emergence of pathogens that were suppressed during the pandemic, immune vaccination, and the resumption of social activities, suggesting the need to establish an age- and pathogen-specific surveillance system for future infec-tious disease responses.

ObjectiveTo assess the association between serum galectin-3 levels and echocardiographic changes in hospitalized patients with coronavirus disease 2019 across different disease severities and stages, including the post- coronavirus disease 2019 period.MethodsThis retrospective observational study included 86 adults with reverse transcription polymerase chain reaction-confirmed coronavirus disease 2019 who were treated at the First University Clinic of Tbilisi State Medical University. Patients were divided into four groups: moderate disease within the first week of symptom onset (n = 21), severe disease during weeks 2-3 (n = 20), critical illness not requiring mechanical ventilation (n = 20), and post- coronavirus disease 2019 status 2 months after infection (n = 25). Serum galectin-3 was measured using an enzyme-linked immunosorbent assay, and transthoracic echocardiography was performed in all patients. Clinical, laboratory, and echocardiographic variables were compared across groups, and associations with galectin-3 were assessed using Spearman rank correlations. Group comparisons were performed using independent-samples t tests and one-way analysis of variance. Because normality and homogeneity of variance were not formally assessed and no adjustment for multiple testing was applied, the findings were interpreted as exploratory.ResultsGalectin-3 levels increased across the acute disease groups and were highest in critically ill nonventilated patients, whereas lower levels were observed in the post- coronavirus disease 2019 group. Nominally significant between-group differences were observed in ascending aortic diameter, left ventricular volumes and mass, left atrial diameter, pulmonary artery diameter, and selected right heart functional parameters. Galectin-3 showed nominally significant correlations with atrial dimensions, right ventricular fractional area change, and selected left ventricular structural parameters, although the pattern varied across groups.ConclusionsSerum galectin-3 was associated with selected echocardiographic abnormalities, with patterns varying according to coronavirus disease 2019 severity and disease stage. Combined assessment of galectin-3 and echocardiographic findings may help characterize cardiovascular involvement. However, because of the small subgroup sizes, unadjusted multiple comparisons, unassessed assumptions of parametric tests, and exclusion of mechanically ventilated patients, these findings are hypothesis-generating and should be interpreted with caution.

This mixed-methods study assessed whether reasoning-enabled large language models (LLMs) can classify stances towards COVID-19 vaccination on X (formerly Twitter) and whether model-generated chain-of-thought (CoT) summaries contain reasoning failures relevant to transparent and auditable public health applications. Zero-shot stance classification by o4-mini and Gemini 2.5 Flash (Gemini) was evaluated on 3,060 rehydrated COVID-19 vaccination tweets against human-annotated labels (positive, negative, neutral). We reported accuracy and macro-F1, measured CoT availability, and qualitatively analysed dual-error cases (tweets misclassified by both models) using Mayring's content analysis guided by the FUTURE-AI framework. At each model's best-performing setting, both models reached macro-F1 around 0.8, with o4-mini outperforming Gemini (accuracy 0.819 vs. 0.799, McNemar p = 0.0015; Δmacro-F1 = 0.020, 95% CI 0.008-0.032). Under the reasoning-intensive settings, CoT availability differed: Gemini returned a reasoning summary for all tweets, whereas o4-mini did so for 64.7%. Among 1,981 tweets with CoTs from both models, 295 (14.9%) were dual-errors; in 88.8%, both models produced the same wrong label, suggesting shared failure modes. Qualitatively, both models showed the same errors: target confusion (policy vs. vaccine), literal readings of sarcasm, and label-rationale mismatches, recurring across models despite their markedly different CoT lengths. Reasoning LLMs can therefore classify stance accurately, but their readiness for transparent public health applications depends on whether a CoT is available at all and whether it is coherent with the label it accompanies (label-rationale coherence). CoT availability, label-rationale coherence, and safeguards against systematic reasoning failures offer candidate explainability-readiness metrics, alongside accuracy, for trustworthy digital epidemiology.

Advanced cytokine-based frameworks for pediatric acute respiratory distress syndrome (PARDS) are not readily applicable in routine practice. We investigated whether routinely available admission biomarkers could define inflammatory phenotypes in young children hospitalized with community-acquired pneumonia (CAP) and provide prognostic information for severe CAP and in-hospital PARDS, with potential relevance for resource-constrained settings. In this prospective cohort of children aged < 5 years hospitalized with CAP, seven admission biomarkers (neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, systemic immune-inflammation index, systemic inflammation response index, C-reactive protein, and procalcitonin) were log1p-transformed, standardized, and analyzed using K-means clustering. Associations with severe CAP at admission and definite in-hospital PARDS were examined using age- and sex-adjusted logistic regression. Incremental prognostic performance was evaluated using nested models, receiver operating characteristic analysis, calibration, and decision curve analysis. Among 129 children, three biomarker phenotypes were identified: low-inflammatory (n = 40), intermediate-inflammatory (n = 60), and hyper-inflammatory (n = 29). Compared with the low-inflammatory phenotype, adjusted odds of PARDS were higher in the intermediate-inflammatory (OR 4.00, 95% CI 1.17-18.6; p = 0.043) and hyper-inflammatory phenotypes (OR 11.6, 95% CI 2.90-60.9; p = 0.001). Only the hyper-inflammatory phenotype was independently associated with severe CAP at admission (OR 4.79, 95% CI 1.54-16.2; p = 0.009). Adding biomarker phenotype to age and sex improved PARDS discrimination (AUC 0.571 to 0.716), but discrimination was similar to NLR alone (AUC 0.711), and gain beyond severe CAP was modest.

Routine admission biomarkers identified inflammatory phenotypes associated with severe CAP and PARDS. This low-cost may be relevant in resource-constrained settings but requires external validation.

• Hyperinflammatory biology is associated with worse outcomes in pediatric CAP and PARDS, but most biomarker frameworks rely on research-oriented cytokine panels. • Routine admission biomarkers such as CRP, PCT, and CBC-derived indices are widely available, but their value for inflammatory phenotyping remains unclear.

• Routine admission biomarkers identified three clinically interpretable inflammatory phenotypes in young children hospitalized with CAP. • The hyper-inflammatory phenotype was associated with severe CAP and PARDS; this pragmatic approach may be relevant in resource-constrained settings, although discrimination was similar to NLR alone.

Early identification of patients at high-risk for exacerbation in mild-to-moderate chronic obstructive pulmonary disease (COPD), particularly those with chronic bronchitis (CB) phenotype exhibiting worse outcomes, is critical for slowing lung function decline and reducing the subsequent healthcare burden. While the soluble receptor for advanced glycation end products (sRAGE) correlates with emphysema severity in COPD, in mild-to-moderate COPD or patients with CB symptoms, no prospective investigation has evaluated the association between sRAGE and exacerbation risk.

Sub-cohort data from the COMPASS study, a prospective study in China, were analyzed. Associations between sRAGE levels and exacerbations were assessed treating sRAGE as a continuous variable and categorized into tertile. Negative binomial regression and cox regression were used to assess exacerbation rate and onset time. Subsequent analyses of lung function, COPD Assessment Tool (CAT) score and high-resolution computed tomography (HRCT) parameters were evaluated using linear mixed-effects models based on a cutoff determined by Akaike information criteria (AIC).

In the COMPASS sub-cohort, 201 patients had mild-to-moderate COPD and 230 individuals reported CB symptoms. Continuous sRAGE level was not associated with exacerbation risks, but with earlier onset time. In addition, participants in the lowest sRAGE tertile exhibited significant shorter time to first exacerbation in both populations. The lowest tertile was also associated with increased exacerbation risks in patients with CB and showed a borderline significant association in patients with mild-to-moderate COPD. No significant differences in lung function decline, HRCT changes, or CAT score progression were observed between the low- and high-sRAGE groups.

Lower sRAGE level may be associated with an increased exacerbation risk in mild-to-moderate COPD and in patients with CB symptoms. Blood sRAGE is a candidate biomarker for identifying subgroups at higher exacerbation risk in populations at high risk of disease progression.

Clinicaltrials.gov identifier NCT04853225; GSK study code 208630.

Venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), significantly contributes to morbidity and mortality among hospitalized patients. Despite the existence of various VTE risk assessment models (RAMs), their performance in accuracy, sensitivity and specificity remain suboptimal, highlighting opportunities to improve predictive accuracy for clinical decision-making.

We conducted a retrospective multicenter study involving three hospitals, which enrolled patients with VTE from January 1, 2021, to December 30, 2023. A novel RAM (Weng score) was developed through three different strategies: clinical knowledge-driven model (Model A), data-driven model (Model B), and decision tree-based model (Model C). The primary outcome was in-hospital VTE. Prediction of PE alone was examined as a secondary outcome. Model performance was evaluated through discrimination, calibration, precision, and decision curve analysis (DCA).

A total of 1,791 patients were analyzed, with 680 VTE events recorded during hospitalization. The Weng score, derived from Model A, demonstrated superior predictive performance for VTE and PE compared to existing RAMs, with an area under the receiver operating characteristic curve (AUROC) of 0.895 (95% confidence interval (CI) [0.880-0.909]) for VTE and 0.877 (95% CI [0.851-0.903]) for PE. In comparison, the AUROCs for existing RAMs (Caprini, Padua, Wells, Geneva, and Autar scores) ranged from 0.687 to 0.789 for VTE prediction and from 0.682 to 0.769 for PE prediction. The Weng score also demonstrated excellent calibration and discrimination, outperforming the Caprini, Padua, Wells, Geneva, and Autar scores in hospitalized patients. The Weng score's clinical utility for relative risk stratification was further supported by DCA within this case-control sampled cohort, showing a higher net benefit in predicting VTE and PE than existing RAMs.

We developed and internally validated the Weng score using retrospective data from Chinese hospitals. While it showed more favorable calibration and discrimination than existing RAMs in our cohort, external validation in diverse settings and prospective studies accounting for anticoagulation management are essential before clinical adoption. The Weng score is intended for VTE risk stratification only; clinical decisions regarding thromboprophylaxis should integrate both VTE and bleeding risk assessments using validated tools. Because the current model was derived from a case-control sampled cohort (oversampled VTE events), the absolute risk estimates and net benefit findings reflect relative risk ranking and require external calibration in a representative prospective cohort before any clinical implementation.