CHEST

Background: During emergency tracheal intubation, clinicians use waveform capnography or colorimetric carbon dioxide (CO2) detection to confirm exhaled CO2 from the endotracheal tube to determine its location. Both devices reliably yield a negative result when the tube is in the esophagus. However, in most emergency intubations the tube is correctly located in the trachea, and in this circumstance the relative diagnostic accuracy of the two devices is uncertain. If one device more often fails to detect exhaled CO2 despite correct tube placement (a false negative result), this could lead to unnecessary tube removal, repeated intubation attempts, and an increased risk of serious complications. Research Question: Among critically ill adults undergoing emergency tracheal intubation, is waveform capnography or colorimetric CO2 detection more sensitive for identifying an endotracheal tube correctly placed in the trachea? Study Design and Methods: The WAVE study is a pragmatic, prospective, paired diagnostic accuracy study conducted in emergency departments and intensive care units in the United States. The study compares waveform capnography and colorimetric CO2 detection (index tests) against a non-CO2-based reference standard for endotracheal tube location among 2,092 critically ill adults undergoing tracheal intubation. The primary outcome is the sensitivity of each index test for identifying a tube located in the trachea. The secondary outcome is specificity for identifying a tube located outside the trachea. Results: Enrollment began on April 21, 2025 and is ongoing, with completion anticipated in 2026. Interpretation: The WAVE study will provide prospective evidence comparing waveform capnography with colorimetric CO2 detection for determination of endotracheal tube location in critically ill adults. These findings may inform clinical practice and guideline recommendations regarding the optimal test of exhaled CO2 for use in the ED and ICU. Prespecifying the protocol and statistical analysis plan before completion of enrollment enhances the rigor, reproducibility, and transparency of the study.

Study Question. COPD exacerbations with respiratory failure have an outsized impact on morbidity, mortality, and healthcare costs. Physicians treat these severe events with high doses of corticosteroids out of concern for intrinsic corticosteroid resistance. This study was designed to determine whether people who suffer from exacerbations with respiratory failure have intrinsic steroid resistance that could affect that decision. Patients and Methods. A matched, case-control study was performed at University of Colorado. Subjects were enrolled into the Exacerbation group if they were hospitalized for an exacerbation requiring ventilation, or into the Stable COPD group if they had no inpatient exacerbation within 12 months. Stable COPD subjects were frequency-matched to Exacerbation subjects for age, smoking status and FEV1%. Corticosteroid sensitivity was assessed ex vivo and in vivo >45 days following the last exacerbation during a period of relative quiescence. The primary outcome was the group difference in methylprednisolone IC50 for LPS-stimulated IL-8 release by mononuclear cells. Measurements and Main Results. The Exacerbation group (N=8) had more exacerbations in the prior year (6.0 vs. 0.5; p<0.01) and had higher baseline blood lymphocytes (p<0.01) versus Stable controls (N=8). Ex vivo, methylprednisolone equally suppressed IL-8, IL-6, TNFalpha and IL-1bets release by mononuclear cells and whole blood across groups. In vivo, methylprednisolone decreased plasma mediators and mononuclear cells in both groups but had a small differential effect on mononuclear cell gene expression. Answer to the Question. Exacerbations with respiratory failure have persistent lymphocytosis, an altered gene response to corticosteroids, but no corticosteroid resistance to suppress inflammation.

Background: Sex differences in asthma are well recognized, but sex-associated airway DNA methylation (DNAm) patterns and their relationships with gene expression remain incompletely characterized. We therefore conducted cis- and trans-expression quantitative trait methylation (eQTM) analyses using sex-stratified and CpG-by-sex interaction models. Methods: We analyzed nasal airway epithelial DNAm and transcriptomic data from 298 participants in the CAAPA discovery cohort. Asthma-associated differentially methylated CpGs (DMCs) were identified using sex-combined and sex-stratified models. Cis- and trans-eQTM analyses were then performed in participants with asthma using sex-combined, sex-stratified, and CpG-by-sex interaction models. To prioritize asthma-associated regulatory relationships, eQTMs that were also significant in non-asthmatic controls were excluded from downstream analyses. Replication in an independent bronchial epithelial cohort was defined by concordant effect direction and P < 0.05. Replicated targets underwent pathway and network analyses. Findings: Asthma-associated DMCs were predominantly hypomethylated in females and hypermethylated in males. Among trans-eQTMs, 379 associations were shared across all four models, whereas the majority were model-specific, indicating extensive sex-dependent regulatory architecture. We replicated 3, 1, 9, and 2 cis-eQTMs and 96, 84, 216, and 41 trans-eQTMs in the sex-combined, female-stratified, male-stratified, and interaction models, respectively. Sex-combined targets implicated innate immune and inflammatory signaling (TNF, IRF8); female-stratified targets implicated cytoskeletal remodeling, focal adhesion, and PI3K-Akt signaling (ITGA10); male-stratified targets implicated transcriptional regulation and post-translational modification (MOV10, HDAC7, CSF1, KAT2A); and interaction targets implicated lipid-mediator metabolism and inflammatory signaling and resolution (CYP2C9, LTB4R, CCL19). Our sex-dependent analyses showed that the majority of eQTMs resided within the trans-regulatory architecture, suggesting that sex-dependent gene expression differences are associated predominantly with trans rather than cis regulatory mechanisms. Pathway analyses revealed enrichment of innate immunity in the sex-combined model, structural remodeling in females, transcriptional control in males, and inflammatory resolution in the sex-interaction model. Interpretation: Sex-specific and sex-dependent airway methylation-expression associations were largely undetected by sex-combined analysis. Sex-aware analytical models may help resolve epigenetic mechanisms underlying sex differences in asthma biology.

Background: Chronic obstructive pulmonary disease (COPD) is characterized by airway remodeling, which may extend to the central airways and alter tracheal morphology. Whether tracheal shape relates to disease severity, quality of life, functional capacity, or survival is not established. Objective: To determine whether CT-derived tracheal lumen eccentricity is associated with cross-sectional and longitudinal clinical outcomes in COPD. Methods: We analyzed 8,383 participants from the multicenter COPDGene study. At each axial cross-section of the trachea, lumen eccentricity ({epsilon}) was computed from the major and minor axis lengths, and the participant-level value was taken as the arithmetic mean across all tracheal slices; {epsilon} = 0 denotes a circular lumen and values approaching 1 denote progressively elongated ellipses. Baseline outcomes were spirometry, 6-minute walk distance (6MWD), St. George's Respiratory Questionnaire (SGRQ) total score, and modified Medical Research Council (mMRC) dyspnea score. Longitudinal outcomes were annualized change in FEV1 and all-cause mortality. Multivariable linear and Cox proportional hazards models were adjusted for [age, sex, race, BMI, smoking status, pack-years, scanner model]. Results: Tracheal eccentricity rose monotonically with severity of airflow obstruction (0.60{+/-}0.07, 0.61{+/-}0.07, 0.62{+/-}0.07, 0.64{+/-}0.08, and 0.64{+/-}0.08 in GOLD stages 0 through 4, respectively). Higher eccentricity was independently associated with lower FEV1 ({beta} = -0.30 L, 95% CI -0.47 to -0.14; p < 0.001), lower FEV1/FVC ({beta} = -0.06, 95% CI -0.09 to -0.04; p < 0.001), shorter 6MWD ({beta} = -15.7 m, 95% CI -25.2 to -6.2; p = 0.001), higher SGRQ score ({beta} = 16.9, 95% CI 11.2 to 22.5; p < 0.001), higher mMRC dyspnea score ({beta} = 0.4, 95% CI 0.1 to 0.6; p = 0.007), and accelerated lung function decline ({beta} = -23.8 mL/year, 95% CI -45.13 to -2.48; p = 0.029). Eccentricity was also associated with all-cause mortality (adjusted HR 1.81, 95% CI 1.01 to 3.26; p = 0.047). Conclusions: Tracheal lumen eccentricity is associated with worse quality of life, reduced functional capacity, accelerated lung function decline, and mortality in COPD, supporting central airway morphology as a candidate imaging biomarker.

Introduction Maintaining long-term positive airway pressure (CPAP) adherence is a major challenge when treating obstructive sleep apnea (OSA). We evaluated factors associated with one-year CPAP adherence and developed models predicting Month-12 adherence and CPAP-use trajectories. Methods CPAP-naive adults with OSA (apnea-hypopnea index[&ge;]5) at Kaiser Permanente Southern California with 364-day observation period were analyzed (development/internal-validation cohort, n=14,906; temporal holdout, n=1,676). Logistic regression evaluated baseline associations with Month-12 adherence (mean[&ge;]4 hours/night during days 331-360). XGBoost models utilized baseline features (clinical, sleep-study, questionnaire data), early CPAP data (days 1-7), or combined to predict long-term CPAP use. A latent-class model identified trajectories over days 8-364. Predictive models were evaluated in both random and temporal holdouts. Results About one-third (34.6%) of the primary cohort (49.7{+/-}13.8 years, 63.2% male) were adherent during Month-12. Higher AHI and longer self-reported sleep duration favored adherence while greater comorbidity burden, depression, and substance-use disorder were associated with lower adherence. In held-out testing, adherence prediction using baseline features yielded an AUROC of 0.678. Using CPAP-only data, AUROCs were higher from 0.787 with seven days of usage data to 0.816 and 0.854 with 14 and 30 days. Models combining CPAP and baseline features increased Day-7 AUROC to 0.810. We identified four one-year trajectories: Early-Decliners (42.9%), Gradual-Decliners (17.4%), Stable-Moderate (19.0%), Stable-High (20.7%). Models predicting patient trajectory revealed macro-average AUROCs of 0.619, 0.756, and 0.770 for baseline, CPAP-only, and combined models. Conclusion Baseline predictors may identify strategies for adherence support, while early CPAP data substantially improved predictive performance. These models may guide personalized follow-up but should not restrict CPAP access. External and implementation validation are needed.

Background and Objective: Lung function reference equations are essential for interpreting spirometry and related modalities, yet existing tools are fragmented, modality- and language-specific (predominantly R), and lack integrated severity classifiers. We present PySpiro, an open-source Python library unifying 23 spirometry, 1 multiple breath washout, 2 diffusing capacity, 1 static lung volume, and 2 oscillometry reference equations under a single API, with 10 clinical severity classifiers. Methods: PySpiro was implemented in Python 3 using NumPy and pandas, with all equations coded directly from literature. The unified API exposes predicted medians, z-scores, limits of normal (LLN/ULN), and raw Lambda-Mu-Sigma (LMS) triplets per parameter, plus a vectorised compute() method for batch processing at the DataFrame level. Unit-tests covering every equation and classifier verify agreement with published values and confirm correct boundary, out-of-range, and missing-input handling. Results: PySpiro covers spirometry equations from 1971 to 2022 across Caucasian, Asian, Indian, and race-neutral global populations, plus non-spirometry modalities (multiple breath washout, diffusing capacity, static lung volumes, oscillometry). Classifiers span COPD, interstitial lung disease, pattern and severity grading, bronchodilator reversibility, and primary ciliary dyskinesia. Compared with rspiro (5 equations, 4 spirometry) and the only prior Python package (spiref, 2 equations, no z-scores or % predicted), PySpiro offers substantially broader coverage, multi-modality support, a richer output API, and 10 classifiers absent from both. Conclusions: PySpiro is freely available (PyPI: pyspiro; doi:10.5281/zenodo.15519193) and provides a comprehensive, validated, and extensible platform for lung function data analysis in research pipelines.

Rationale: Conventional CT-derived pulmonary vascular metrics quantify vascular pruning but do not assess the spatial heterogeneity of vascular architecture. In COPD, peripheral microvascular pruning coexists with compensatory central vessel dilation, paradoxically preserving total vessel volume while reducing effective vascular penetration into lung parenchyma. Objectives: To develop and evaluate Vascular Proximity Coverage (VPC), a distance-based CT biomarker quantifying the spatial efficiency of pulmonary vascular distribution. Methods: This retrospective analysis included 7,881 COPDGene participants [PHASE] with complete vascular segmentations. VPC was estimated by integrating the area under the cumulative lung tissue coverage curve at distances of 0-10 mm from segmented vessel voxels. Multivariable models assessed associations with lung function, six-minute walk distance (6MWD), St. George's Respiratory Questionnaire (SGRQ) score, mMRC dyspnea score, and all-cause mortality over a median follow-up of [10.8] years, adjusting for demographics, smoking, emphysema, scanner type, and BV5. The cohort was stratified at median BV5 and VPC into concordant (high/high, low/low) and discordant (high-BV5/low-VPC, low-BV5/high-VPC) vascular phenotypes. Results: VPC declined across disease severity (5.82 {+/-} 0.50 [UNITS] in non-smokers to 5.21 {+/-} 0.64 in GOLD 4; p < 0.001). Among spirometrically normal smokers, current smokers had lower VPC than former smokers (5.31 vs. 5.68; p < 0.001). Each standard deviation decrease in VPC was independently associated with reduced 6MWD ({beta} = -16.8 m, 95% CI: -19.5 to -14.0), worse SGRQ ({beta} = 1.9, 95% CI: 1.4 to 2.4), and greater dyspnea (RR = 1.05, 95% CI: 1.03 to 1.08; all p < 0.001). The lowest VPC quartile was associated with 38% higher mortality than the reference quartile (HR = 1.38, 95% CI: 1.18 to 1.62; p < 0.001). The high-BV5/low-VPC phenotype showed 24% higher mortality than high-BV5/high-VPC (HR = 1.24, 95% CI: 1.08 to 1.42; p = 0.002) despite preserved vessel volume. Conclusions: VPC captures spatial heterogeneity of vascular architecture independent of vessel volume, is reduced in smokers without airflow obstruction, and identifies excess mortality risk among individuals whose total vessel volume appears preserved.

Background: Hospital readmission is a common and costly outcome in patients with chronic obstructive pulmonary disease (COPD). Multimodal deep learning models combining structured electronic health record (EHR) data with physiological, imaging, and textual data have been proposed to improve its prediction, but whether such fusion outperforms EHR data alone, once temporal-leakage safeguards are applied, remains unclear. Methods: Using a stratified MIMIC-IV cohort of COPD patients enriched to about 50% readmission (natural prevalence 18.1%; n=1,279; 895/192/192 training/validation/test), we predicted 30-day, all-cause readmission from four frozen, pre-trained encoders (electrocardiogram, chest radiograph, clinical notes, structured EHR). We audited and corrected temporal leakage in the admission-linkage pipeline of each modality, then compared three fusion architectures (naive concatenation, scalar gated fusion, EHR-anchored residual fusion) against unimodal baselines and an EHR-alone reference, using bootstrap significance testing, a missingness stress test, and calibration analysis with and without post-hoc recalibration. Results: The audit corrected three previously undetected temporal-alignment issues, affecting 5-9% of patients for clinical notes, 29-43% for chest radiographs, and 53-58% for ECG. After correction, every modality except EHR showed near-chance discrimination alone (ROC-AUC 0.46-0.53), confirmed with a non-linear adapter probe. No fusion architecture significantly outperformed the EHR-alone baseline (ROC-AUC 0.63); removing auxiliary modalities improved rather than degraded fusion performance under stress testing. The raw Brier score was lower for all three fusion architectures than for EHR-alone (about 0.24 versus 0.32), but this reflected overconfidence of the EHR-alone probe (calibration slope 0.12); after logistic recalibration fitted on the validation set, Brier scores were indistinguishable (EHR-alone 0.239; fusion 0.240-0.244; all paired bootstrap 95% CIs included zero). Conclusions: Structured EHR data alone matched the discrimination and, after recalibration, the calibration of more complex multimodal fusion for 30-day COPD readmission prediction. These findings highlight the importance of rigorous temporal-leakage auditing in multimodal clinical prediction research and suggest that reformulating this task around temporal proximity to the outcome, rather than refining the fusion mechanism further, may be more productive for future work.

Background: Patients increasingly consult free artificial intelligence (AI) chatbots about their symptoms, but these models may favour agreement with the patient over accuracy. In obstructive sleep apnoea (OSA), where most cases in the general population remain undiagnosed and diagnosis depends on referral, this could validate symptom denial. We tested whether consumer chatbots maintain guideline-concordant referral when a simulated patient minimises symptoms. Methods: Seven guideline-concordant and standardised OSA vignettes (apnoea-hypopnoea-index, AHI 8-52 events/h; body mass index 28-42 kg/m2) were presented to five free chatbots (GPT-5, Gemini 2.5 Flash, Claude Sonnet 4.6, DeepSeek V4 Flash and Grok-4.3) under neutral and symptom-minimising conditions (10 simulated conversations per model-vignette-condition cell; N=700). Referral maintenance was scored by a structured rubric and every capitulation was human-confirmed. Results: All chatbots maintained referral recommendations in neutral conversations (350/350; 100%). When simulated patients minimised symptoms and resisted referral, maintenance fell to 64.3% (225/350), a reduction of 35.7 percentage points (95% CI 5.5-66.0; p=0.028). Rates ranged from 55.7% to 85.7% across chatbots, but evidence of between-chatbot differences was inconsistent across analyses. Rates differed across clinical vignettes (p=0.001). A variance analysis estimated greater variation between vignettes than between chatbots. In the three driving-risk vignettes, referral recommendations were abandoned in 46-68% of symptom-minimising conversations, pooling all five chatbots within each vignette. Conclusions: Consumer AI chatbots abandoned correct OSA referral recommendations in about one-third of symptom-minimising conversations, with higher observed abandonment in simulated cases involving driving risk. This may delay appropriate assessment and raises patient- and public-safety concerns.

Abstract Background The Simplified Geneva Score (SGS) incorporates two heart-rate thresholds. Multiple heart-rate measurements are commonly recorded during emergency department (ED) assessment. While the original SGS specifies use of mean heart rate, real-world implementation is unlikely to reflect this, and the effect of alternative heart-rate selection methods on score performance is unknown. Methods SCORES was a retrospective within-subject study of consecutive ED patients at a single urban hospital undergoing computed tomography pulmonary angiography (CTPA) for suspected pulmonary embolism (PE). Four SGS variants were calculated using the first, highest, mean, and lowest recorded pre-CTPA heart rate. Outcomes included score output, risk classification, diagnostic performance, reclassification flow, and downstream imaging utilisation. Precision estimates and comparisons were derived using empirical resampling techniques. Results 590 episodes were reviewed with 581 included for analysis. PE was confirmed in 50 cases (8.6%). Mean SGS differed significantly across all heart-rate implementations (p<0.0001). The proportion classified as high risk varied from 15% using the lowest heart rate to 40% using the highest heart rate. Reclassification across the SGS threshold was common, with 25% of episodes changing risk category unidirectionally between lowest and highest heart rate implementations. SGS(lowest HR) compared with SGS(mean HR) and SGS(highest HR) demonstrated higher specificity (87.6% vs 73.4% vs 62.1%), improved AUROC (0.744 vs 0.704 vs 0.659), and reduced downstream CTPA utilisation (818 vs 876 vs 920 scans per 100 PEs diagnosed), with no additional pulmonary emboli missed within the complete diagnostic pathway. Conclusions Heart-rate selection materially influences SGS output, risk classification, diagnostic performance, and imaging utilisation. In this retrospective cohort, use of the lowest recorded heart rate produced the most favourable overall balance of discrimination and diagnostic efficiency, supporting prospective validation.

Aims. It is unknown whether large language models (LLMs) can predict a trial's result at the design stage, from the information available when it is registered. We tested three LLMs prospectively on the Late-Breaking Science programme of the European Society of Cardiology (ESC) Congress 2026. Methods and results. Of 139 trials, 50 were selected by pre-specified criteria, and predictions were locked and registered on the Open Science Framework before the congress. Claude Fable 5, GPT-5.5 and Gemini 3.5 Flash, without web access, received each title and an investigator-compiled summary of the registered design. They gave the probability that the primary endpoint would be met, and a point estimate with an 80% prediction interval for a locked effect measure. Results were adjudicated from publications (19) or presenters' slides (31). Of 48 scorable trials, 28 (58%) were positive. The area under the receiver operating characteristic curve was 0.80 (95% CI 0.66-0.92) for Claude, 0.79 (0.66-0.91) for GPT and 0.74 (0.59-0.87) for Gemini, with accuracies of 73%, 71% and 67%. Among 26 trials whose effect was reported exactly as locked, the direction was correct in 88% of predictions and the median multiplicative error of ratio estimates was about 15%. The 80% prediction intervals contained the observed effect in 73.1% (Claude), 84.6% (GPT) and 57.7% (Gemini). Conclusion. From the registered design alone, LLMs showed statistically significant discrimination between positive and negative trials and usually predicted the direction and approximate size of the effect. Such predictions might help trial design, and further studies are needed.

Early risk stratification of COVID-19 patients at admission informs triage and resource allocation, particularly in low- and middle-income settings where intensive-care capacity is constrained. Most published models are not interpretable, rarely externally validated, and developed almost exclusively in high-income cohorts. We developed an interpretable random-forest classifier for three-class COVID-19 severity (Mild: not ventilated, survived; Severe: ventilated, survived; Fatal: died) using only data available within 24 hours of admission in 321 hospitalised PCR-confirmed COVID-19 patients in Pakistan. From 2,278 raw variables we retained 20 predictors by clinical screening and permutation-importance selection, with leakage-controlled multiple imputation by chained equations and SHapley Additive exPlanations (SHAP) for interpretation. A 15-feature variant was evaluated in two independent public cohorts: the Chinese iCTCF cohort (n = 894) and an Italian complete-blood-count cohort (n = 1,218). The development model achieved a macro F1 of 0.55, identifying the Fatal class most reliably (F1 0.67, recall 0.73) and Mild cases least well. SHAP identified a parsimonious set of admission-time predictors (oxygen saturation, lactate dehydrogenase, urea, C-reactive protein, age and radiographic pneumonia), including a raised urea-to-creatinine ratio in fatal cases most consistent with dehydration (median urea 80 vs 50 mg/dL, fatal vs mild; p < 0.001). Externally, the model retained useful discrimination in the Chinese cohort (mortality AUC 0.82 in 719 patients with recorded outcomes; severe-or-worse AUC 0.75; macro F1 0.50) but only modest discrimination in the Italian cohort (AUC 0.62, 95% CI 0.56-0.68), which we treat as a feature-availability stress test rather than a second validation. Cross-cohort SHAP rankings correlated strongly in China (Spearman {rho} = 0.91) and moderately in Italy ({rho} = 0.76), where most laboratory predictors were imputed and carried no information beyond the complete blood count; discrimination tracked how many predictors were actually measured. The admission-time signature converges with established markers of hypoperfusion and inflammation.

Background. Polygenic risk scores (PRS) capture inherited susceptibility to asthma, whereas methylation risk scores (MRS) may reflect disease-associated epigenetic variation. We assessed whether these scores differ in patients requiring biologic therapy and whether MRS change after mepolizumab initiation. Methods. Cross-sectional analyses in the GEMAS study included 699 participants with genotype data (428 with asthma, 271 controls) and 668 with blood DNA methylation data (408 with asthma, 260 controls). We evaluated 35 existing asthma PRS and derived six MRS from published epigenomic asthma studies. Longitudinal MRS changes after mepolizumab initiation were assessed in 16 GEMAS patients and were replicated in 20 patients from the MEGA study. Results. Twenty-eight PRS differentiated biologic-treated patients from controls (areas under the curve [AUC], 0.68-0.75), but none distinguished biologic-treated and non-treated patients after multiple-testing correction. MRS was higher in asthma without biologic than in controls (odds ratio [OR], 1.64; 95% CI, 1.34-1.99) and lower in biologic-treated than in asthma without biologic patients (OR, 0.54; 95% CI, 0.35-0.81). MRS did not differ significantly between biologic-treated patients and controls. Following mepolizumab initiation, MRS decreased at 6 and 12 months in GEMAS (p=4.81x10-5 and p=2.40x10-5, respectively). The 6-month decrease was also observed in MEGA (p=3.25x10-9). MRS epigenetic markers were predominantly hypermethylated, opposite to their described pattern in asthma risk. Conclusions. Asthma PRS did not differ significantly between biologic-treated and non-biologic-treated patients, but asthma MRS were lower in biologic-treated patients. The MRS decreased after mepolizumab initiation in two independent cohorts, consistent with a reversal of asthma-associated epigenetic patterns.

INTRODUCTION Large language models (LLMs) have demonstrated substantial performance on medical licensing and board examinations, but their application to anesthesiology remains less well studied. Prior evaluations have not compared current-generation models from different developers or assessed performance on figure-dependent questions that earlier models lacked the capability to interpret. This study evaluates the performance of two current-generation LLMs on a comprehensive anesthesiology in-training examination (ITE) review question bank, including figure-dependent items. METHODS A total of 1,001 single-best-answer questions from an anesthesiology ITE review text, spanning 11 content chapters, were administered to Claude Opus 5 and GPT-5.6. No questions were excluded. Each item was presented once in a fresh, stateless context with no tool access, retrieval, answer key, or explanation provided in the prompt. Performance was analyzed by question format, and all 27 figure-dependent items were administered with their published figures supplied. Accuracy between models was compared using McNemar's test. RESULTS Claude Opus 5 answered 947/1,001 items correctly (94.6%; 95% CI, 93.0-95.8), and GPT-5.6 answered 946/1,001 correctly (94.5%; 95% CI, 92.9-95.8); the difference was not significant (McNemar p=1.00). Accuracy was similar across most question formats. On standard questions, accuracy was 94.3% and 94.6% for Claude Opus 5 and GPT-5.6, respectively; on the 18 figure-based questions with figures supplied, accuracy was 94.4% and 83.3%, respectively; and both models achieved 100% accuracy on image-option items. Withholding figures from the same 18 figure-based questions reduced pooled accuracy from 88.9% to 52.8% (exact McNemar p=0.03 and p=0.04 for Claude Opus 5 and GPT-5.6, respectively). The models agreed on 957/1,001 items (95.6%). Of the 33 items both models answered incorrectly, they selected the same incorrect option on 32 (97%). DISCUSSION Current-generation LLMs achieved approximately 95% accuracy, substantially improving on our prior results with earlier-generation models and placing both models within the highest band of the ITE normative framework, corresponding approximately to the 99th percentile across training levels. By comparison, our prior work with previous models placed ChatGPT-3.5 at the 52nd, 3rd, and 1st percentiles and ChatGPT-4.0 at the 99th, 95th, and 84th percentiles across increasing levels of training. Multimodal capability now permits successful interpretation of many figure-dependent questions, although performance declines markedly when required visual information is unavailable, and highly concordant errors remain. These findings support an increasingly promising role for LLMs as accessible, on-demand educational adjuncts for anesthesiology trainees, provided complete source material is supplied and outputs are critically reviewed.

Objective To evaluate the information quality, reliability, readability, and content deficits of ChatGPT, DeepSeek, and Doubao in answering frequently asked questions about bee stings. Methods Twenty-five high-frequency patient questions were selected based on Google Trends, domestic health platforms, and allergy guidelines, and submitted to the three LLMs, yielding 75 responses. Two emergency physicians independently assessed quality using GQS, mDISCERN, QAMAI, EQIP, and PEMAT-U/A. Readability was evaluated using six indices (ARI, FRES, GFI, FKGL, CLI, SMOG). Content deficits were classified qualitatively. Twenty emergency department patients evaluated the comprehensibility and credibility of AI responses. Results Significant differences were found among the three models on all quality metrics (all P < 0.01). DeepSeek achieved the highest mDISCERN score (median 21.00); ChatGPT led on QAMAI (25.00) and EQIP (75.00%); Doubao ranked lowest on all quality indicators but highest on PEMAT-U (73.00%). No model met the sixth-grade reading level standard. Eleven content deficits were identified: outdated medical advice (45.45%), omissions (36.36%), and hallucinations (18.18%). Doubao received the highest patient preference (45.0%). Conclusion The three LLMs exhibited distinct strengths: DeepSeek in source reliability, ChatGPT in overall quality, and Doubao in understandability. All models had excessively high readability levels, and outdated medical advice was the predominant content deficit. Objective To evaluate the information quality, reliability, readability, and content deficits of ChatGPT, DeepSeek, and Doubao in answering frequently asked questions about bee stings. Methods Twenty-five high-frequency patient questions were selected based on Google Trends, domestic health platforms, and allergy guidelines, and submitted to the three LLMs, yielding 75 responses. Two emergency physicians independently assessed quality using GQS, mDISCERN, QAMAI, EQIP, and PEMAT-U/A. Readability was evaluated using six indices (ARI, FRES, GFI, FKGL, CLI, SMOG). Content deficits were classified qualitatively. Twenty emergency department patients evaluated the comprehensibility and credibility of AI responses. Results Significant differences were found among the three models on all quality metrics (all P < 0.01). DeepSeek achieved the highest mDISCERN score (median 21.00); ChatGPT led on QAMAI (25.00) and EQIP (75.00%); Doubao ranked lowest on all quality indicators but highest on PEMAT-U (73.00%). No model met the sixth-grade reading level standard. Eleven content deficits were identified: outdated medical advice (45.45%), omissions (36.36%), and hallucinations (18.18%). Doubao received the highest patient preference (45.0%). Conclusion The three LLMs exhibited distinct strengths: DeepSeek in source reliability, ChatGPT in overall quality, and Doubao in understandability. All models had excessively high readability levels, and outdated medical advice was the predominant content deficit.

Objectives: Health professions students often receive limited training on how to adjust their communication for patients whose first language is not English (LX), despite the risks for diagnosis, treatment, and patient trust. Based on Communication Accommodation Theory, we developed a brief module to improve students' awareness and knowledge of strategic communication adjustments for LX patients. Methods: We tested feasibility in a single-group, pre-post pilot study guided by the Kirkpatrick framework with undergraduate (n=22) and graduate (n=41) speech-language pathology and psychology students in Massachusetts and North Carolina. The module included instructional content, video and audio examples, interactive activities, and a podcast. Learning was assessed with knowledge checks, and self-reported facilitative and resistant accommodation with the Communication Accommodation Behavior Scale (CABS). After training, students rated their confidence and likelihood of using accommodation strategies. Results: At both sites, facilitative accommodation increased and resistant accommodation decreased significantly, with a larger decrease at the undergraduate site. Knowledge-check accuracy was high at both time points, suggesting a ceiling effect. Students reported moderate- to-high confidence in adapting their speech and high likelihood of using a learned strategy in future clinical encounters. Conclusions: A brief module can be feasibly delivered across institutions and is associated with short-term improvements in self-reported communication accommodation. Practice implications: We recommend future work evaluate whether these changes translate into clinical behavior long-term and refine knowledge-check items to better capture learning gains.

BACKGROUND: As artificial intelligence (AI) enters clinical practice, concerns about deskilling and never-skilling make the independent assessment of clinical competence increasingly important. The AI-Standardized Clinical Examination (ASCE) uses AI to simulate and score standardized clinical encounters, potentially overcoming some of the logistical and standardization constraints of conventional objective structured clinical examinations (OSCEs). Whether it provides valid and reliable assessment at scale is unknown. METHODS: In this prospective multicenter cohort study, final-year medical students from six French medical schools completed a 10-station ASCE six weeks before the national OSCE session. GPT-4o simulated standardized clinical partners; transcripts were subsequently scored by GPT-4.1 against faculty-defined rubrics. The primary outcome was the correlation between ASCE and national OSCE scores. Secondary outcomes were inter-rater reliability of expert transcript scoring, alignment of ASCE scores with expert ratings, and internal consistency. RESULTS: Among 948 enrolled students, 588 had paired ASCE and national OSCE scores. ASCE scores were correlated with national OSCE performance (Spearman's {rho}, 0.57; 95% CI, 0.51 to 0.63), with less than 1% of residual variance attributable to center (intraclass correlation coefficient, 0.01). Eight experts scored 20 randomly selected ASCE transcripts; Krippendorff's was 0.77 overall, 0.80 for clinical skills, and 0.70 for communication and attitude skills. Adding GPT-4.1 as an evaluator reduced by 0.02, and its scores did not differ significantly from those of individual experts. Internal consistency was 0.67 across 10 stations and was projected to approach 0.80 with 20 stations. CONCLUSIONS: Scores from an examination delivered and scored by AI were associated with performance on a national high-stakes OSCE, with minimal center-level variation, tentative agreement among expert raters, and automated scores that aligned with expert ratings. These results support AI delivery and scoring of clinical examinations, with expert review of automated scores.

BackgroundMacitentan is a dual endothelin receptor antagonist used in pulmonary arterial hypertension (PAH) and has also been studied in other forms of pulmonary hypertension (PH). Evidence across PH phenotypes has been heterogeneous. This study aims to evaluate the efficacy and safety of macitentan compared with concurrent control therapies across PH phenotypes and to explore whether treatment effects differ according to PH classification. MethodsPubMed, Embase, and the Cochrane database were searched from inception through June 2026. Randomized and non-randomized comparative studies of patients with PH receiving macitentan were eligible. Prespecified efficacy outcomes were 6-minute walk distance (6MWD), pulmonary vascular resistance (PVR), mean pulmonary arterial pressure (mPAP), and cardiac index; prespecified safety outcomes included mortality and adverse events. Random-effects meta-analyses were performed using mean differences (MDs) for continuous outcomes and risk ratios (RRs) for dichotomous outcomes. ResultsFourteen studies met the eligibility criteria. Macitentan was associated with a greater reduction in PVR (MD, -134.50 dyn.s.cm; 95% CI, -226.10 to -42.86) and an increase in cardiac index (MD, 0.44 L/min/m; 95% CI, 0.22 to 0.67). No statistically significant difference was observed for 6MWD (MD, 5.04 m; 95% CI, -15.06 to 25.14) or mPAP (MD, -2.70 mmHg; 95% CI, -6.46 to 1.05). In pooled safety analyses, mortality was lower and risks of anemia, peripheral edema, and discontinuation due to adverse events were higher with macitentan. Heterogeneity and the small number of studies within several phenotype-specific analyses limited the precision of subgroup findings. ConclusionsMacitentan-containing therapy was associated with favorable changes in selected hemodynamic outcomes but not with a statistically significant improvement in 6MWD or mPAP. The safety findings were characterized by higher risks of anemia, peripheral edema, and treatment discontinuation. These results should be interpreted in the context of differences in PH phenotype, dose, comparator, study design, and outcome measurement.

Objectives: To investigate whether the PneumoRator wearable sensor could accurately assess Respiratory Rate (RR) and signs of air hunger (SAH) compared with clinician-observed assessment in adults with and without breathing pattern disorder (BrPD). Design: A pilot cross-sectional observational study. Setting: A research laboratory on a university campus. Participants: Healthy adults aged 18-40, who were not pregnant, no smokers or history of cardiovascular, respiratory or neurological illness. Thirty consented to the study, Twenty-nine had full data for analysis. Interventions: Participants wore the PneumoRator chest-worn sensor continuously for 10 minutes while seated at rest. Main outcome measures: Respiratory Rate was compared between PneumoRator-derived measurements and randomised 1-minute observer manual counts. Clinician-observed SAH and BrPD were assessed using the Breathing Pattern Assessment Tool (BPAT). Participants also completed the Nijmegen Questionnaire. PneumoRator signals were analysed to identify large-volume breaths as potential SAH events. Results: Thirty participants were recruited. Twelve participants met BPAT criteria for BrPD and four met Nijmegen Questionnaire criteria. There was no significant difference between manual and PneumoRator RR measurements (p=0.89). No significant difference was observed between methods for SAH detection (p=0.17). Participants with clinician-observed SAH demonstrated significantly higher PneumoRator signal amplitudes than those without SAH (p=0.035). Conclusion: The PneumoRator produced RR measurements comparable to clinician assessment and demonstrated potential for identifying signs of air hunger. Findings from this pilot study indicate that the PneumoRator device shows promise as a tool for continuous RR monitoring and as an adjunct to BrPD assessment. Larger clinical validation studies are now warranted.

Background: Obesity is associated with airway disease, but the contribution of specific thoracic adipose depots remains unclear. We examined whether intrathoracic adipose tissue (ITAT), defined as adipose tissue within the thoracic cavity excluding pericardial adipose tissue (PAT), was associated with airway phenotypes. Methods: We included Framingham Heart Study participants with CT-derived thoracic adipose tissue measures. Cross-sectional associations of ITAT with CT-based airway wall and lumen measurements, spirometry, and respiratory symptoms were evaluated using multivariable regression. Primary analyses used sex-specific ITAT z-scores and adjusted for waist circumference to account for central obesity. Secondary analyses examined sex-specific ITAT quintiles and compared ITAT and PAT in individual-depot and mutually adjusted models. Results: Among 2,825 adult participants (52.0% female), after adjustment for waist circumference, each 1-SD higher ITAT, relative to the sex-specific distribution, was associated with greater segmental ({beta} [95% CI]: 0.010 mm [0.006, 0.014]) and subsegmental airway wall thickness (0.006 mm [0.002, 0.009]), narrower segmental airway lumen diameter (-0.049 [-0.088, -0.011]), lower FEV1 percent predicted (-1.92 percentage points [-2.72, -1.12]), and higher odds of wheezing (OR [95% CI]: 1.16 [1.02, 1.31]) and dyspnea (1.22 [1.05, 1.42]). Associations were similar after further adjustment for PAT and across ITAT quintiles. In contrast, PAT was not independently associated with airway structural measures after adjustment for ITAT. Conclusion: Higher ITAT volume was associated with greater airway wall thickness and narrower airway lumen, as well as lower FEV1 and greater respiratory symptoms, independent of central obesity and pericardial adiposity. These findings support a depot-specific association of ITAT with airway disease.