International guidelines have emphasized the necessity of evaluating the temporal stability of acute respiratory distress syndrome (ARDS) subphenotypes. This study aimed to assess the temporal stability of subphenotypes of ARDS over 28 days. A reanalysis of a randomized trial was conducted, including patients with COVID-19-related moderate-to-severe ARDS across 43 centers. A K-means clustering was conducted to identify subphenotypes at 7-day intervals from inclusion to day 28. A Bayesian discrete-time Markov model was constructed to assess the temporal stability of subphenotypes. Two subphenotypes were identified among 146 patients. At inclusion, 121 (83%) patients were in the hypoinflammatory subphenotype and 25 (17%) in the hyperinflammatory subphenotype. The hyperinflammatory subphenotype was associated with higher rates of organ failure, higher plasma levels of cytokines, chemokines, adhesion molecules, and proangiogenic factors, and lower endothelial stability than the hypoinflammatory subphenotype. The hyperinflammatory subphenotype was associated with higher 28-day mortality (13/25, 52% vs. 30/121, 25%, p = 0.001) and fewer ventilatory-free-days through day 28 (p < 0.01) than the hypoinflammatory subphenotype. In the Bayesian Markov model, over 7-day intervals, patients in the hypoinflammatory subphenotype had a higher probability of remaining hypoinflammatory (70%) or being extubated (17%) than of progressing to the hyperinflammatory subphenotype (7%). Inversely, patients in the hyperinflammatory subphenotype had a higher probability of remaining in the hyperinflammatory subphenotype (52%) or dying (23%) than of transitioning to the hypoinflammatory subphenotype (20%) or being extubated (5%). Inflammatory subphenotypes were stable in COVID-19-related ARDS, with few transitions over 28 days. Monitoring these subphenotypes could be valuable for assessing patient trajectories and treatment responses.

Introduction: Chronic hypercapnic respiratory failure often necessitates non-invasive ventilation (NIV) at home. Our study aimed to assess the static and dynamic performance of six modern NIV home ventilators and one intensive care unit (ICU) ventilator and to identify performance clusters among the devices. Methods: A two-compartment lung model was connected to seven NIV ventilators (Sefam Stan, Philips A40, Philips DreamStation, Resmed Lumis 150, Löwenstein PrismaVent 30C, Löwenstein PrismaVent 40 and BellaVista 1000) in pressure-support mode. Static and dynamic (triggering and pressurisation) performances were assessed through three distinct clinical phenotypes and four levels of unintentional leak. Clustering analysis was performed using K-means. Results: For each of the seven ventilators, 144 conditions were tested, and a total of 3024 cycles were analysed. Static and dynamic performances were good to excellent across home ventilators, significantly higher than the ICU ventilator. Clustering analysis identified three performance clusters. Cluster 1 (Sefam Stan and Philips A40) showed significantly more precise accuracy of inspiratory pressure than Cluster 2 (Philips DreamStation, Resmed Lumis 150, Löwenstein PrismaVent 30C and Löwenstein PrismaVent 40) and Cluster 3 (BellaVista 1000): mean error=4.3%±5.1% versus 8.5%±6.7% versus 10.6%±14.7% respectively, p<0.001. For the triggering delay, Cluster 1 displayed shorter delays than Cluster 2 and Cluster 3 (41±5 ms vs 58±11 ms vs 67±13 ms, respectively, p<0.001). For the pressurisation delay, Cluster 1 displayed shorter delays than Cluster 2 and Cluster 3 (42±6 ms vs 64±14 ms vs 87±14 ms, respectively, p<0.001). For the pressure-time product at 300 ms, Cluster 1 displayed higher area under the curve for the first 300 ms than Cluster 2 and Cluster 3 (2.1±1.1 cmH ₂ O /s vs 1.6±0.8 cmH ₂ O/s vs 1.3±1.0 cmH ₂ O/s, respectively, p<0.001). Continuous unintentional leaks did not modify the pressurisation performances in Cluster 1 but altered them in Clusters 2 and 3. Conclusion: The six NIV home ventilators demonstrated superior performance compared with the tested ICU ventilator. The ventilators of Cluster 1 were identified as top performers in clustering analysis and compensated for unintentional continuous leaks.

The latest reform of French medical studies has moved the National Ranking Examination before residency to the beginning of the sixth-year for undergraduate medical students, thus placing unprecedented workload during the preceding summer. The main objective was to determine whether study conditions and psychosocial factors were associated with student success in this model of intense workload. An online survey designed with six student-partners was sent at a French Medical School after the examination in 2023. The primary outcome was student success in achieving their main goal (Ranking, Knowledge, Well-being). A machine-learning model (eXtreme Gradient Boosting) was developed and explained using Artificial Intelligence. An AI-guided multivariate logistic regression was performed, Odd Ratios were calculated. Out of 123 responses, 75 (61%) of the students achieved their main goal. Motivation and socialization during the summer were the two most important variables for predicting student success. In guided multivariate logistic regression, summer motivation (Odd Ratio = 4.12, 95%CI[1.75-10.30]), summer loneliness (Odd Ratio = 0.35, 95%CI[0.14-0.86]), and student's main goal (Ranking, Odd Ratio = 2.94, 95%CI[1.15-7.79]) were associated with student success. Motivation and socialization during the summer preceding high-stakes examinations are strongly predictive of undergraduate medical students' success. This study highlights the importance of well-being during summer for student success.[Box: see text].