Daily Ards Research Analysis
Machine learning-derived subphenotypes in ARDS failed to predict prone-positioning responsiveness despite distinct mortality differences, underscoring unmet needs in multimodal phenotyping. In neonates with spontaneous pneumothorax, nitrogen washout shortened time to radiographic resolution versus targeted oxygen therapy. Bedside lung ultrasound tracked ventilation-related changes in critically ill children, supporting its role in ICU monitoring and weaning.
Summary
Machine learning-derived subphenotypes in ARDS failed to predict prone-positioning responsiveness despite distinct mortality differences, underscoring unmet needs in multimodal phenotyping. In neonates with spontaneous pneumothorax, nitrogen washout shortened time to radiographic resolution versus targeted oxygen therapy. Bedside lung ultrasound tracked ventilation-related changes in critically ill children, supporting its role in ICU monitoring and weaning.
Research Themes
- ARDS heterogeneity and prone positioning subphenotypes
- Neonatal spontaneous pneumothorax management strategies
- Point-of-care lung ultrasound in pediatric mechanical ventilation
Selected Articles
1. Subphenotyping prone position responders with machine learning.
Unsupervised machine learning on 353 mechanically ventilated ARDS patients identified three subphenotypes with different mortality, but could not predict which patients would benefit from prone positioning. The study highlights ARDS heterogeneity and the limitations of current clinical variables for precision stratification.
Impact: Provides a data-driven framework for ARDS subphenotyping around prone positioning and clarifies current predictive gaps, guiding future precision ventilation research.
Clinical Implications: Clinicians should continue evidence-based criteria for proning while recognizing heterogeneity; current routine variables may be insufficient to select responders, supporting collection of richer physiologic and multimodal data.
Key Findings
- Three unsupervised machine learning-derived ARDS subphenotypes were identified among 353 proned patients.
- Subphenotypes exhibited different 28-day mortality rates, indicating prognostic heterogeneity.
- Available supine respiratory mechanics and oxygenation variables could not predict benefit from prone positioning.
Methodological Strengths
- Unsupervised clustering across multiple physiological domains (mechanics and oxygenation).
- Clinically relevant primary outcome (28-day mortality) in a sizeable cohort.
Limitations
- Retrospective design with potential selection and information biases.
- Lack of multimodal data (e.g., imaging, biomarkers) may limit predictive performance.
Future Directions: Integrate multimodal data (omics, imaging, ventilator waveform analytics) and prospective validation to enable actionable phenotypes for personalized prone positioning.
BACKGROUND: Acute respiratory distress syndrome (ARDS) is a heterogeneous condition with varying response to prone positioning. We aimed to identify subphenotypes of ARDS patients undergoing prone positioning using machine learning and assess their association with mortality and response to prone positioning. METHODS: In this retrospective observational study, we enrolled 353 mechanically ventilated ARDS patients who underwent at least one prone positioning cycle. Unsupervised machine learning was used to identify subphenotypes based on respiratory mechanics, oxygenation parameters, and demographic variables collected in supine position. The primary outcome was 28-day mortality. Secondary outcomes included response to prone positioning in terms of respiratory system compliance, driving pressure, PaO RESULTS: Three distinct subphenotypes were identified. Cluster 1 (22.9% of whole cohort) had a higher PaO CONCLUSIONS: Distinct ARDS subphenotypes with varying mortality were identified in patients undergoing prone positioning; however, predicting which patients benefited from this intervention based on available data was not possible. These findings underscore the need for continued efforts in phenotyping ARDS through multimodal data to better understand the heterogeneity of this population.
2. Outcomes of spontaneous pneumothorax in neonates: treatments vs. expectant management.
In neonates (≥34 weeks) with spontaneous pneumothorax and respiratory distress/desaturation, nitrogen washout shortened time to resolution versus targeted oxygen therapy (median 31 vs 81 hours; p=0.012). Expectant management was associated with delayed feeding initiation, while time to resolution and length of stay were similar.
Impact: Addresses a long-debated intervention in neonatal pneumothorax with comparative outcomes, providing practice-informing evidence.
Clinical Implications: For symptomatic neonatal spontaneous pneumothorax with desaturation, consider nitrogen washout to expedite resolution; monitor feeding readiness in expectant management pathways.
Key Findings
- Nitrogen washout reduced median time to resolution versus targeted oxygen therapy (31 vs 81 hours; p=0.012) in distressed/desaturating neonates (n=64).
- Expectantly managed distressed neonates (n=87) had delayed initiation of feeds, with similar time to resolution and length of stay.
- Targeted oxygen therapy aimed to maintain saturation ≥93%, serving as a comparator to nitrogen washout.
Methodological Strengths
- Comparative analysis across multiple management strategies (nitrogen washout, targeted oxygen, expectant).
- Clinically meaningful endpoints (time to resolution, feeding initiation, hospital stay).
Limitations
- Retrospective design and non-randomized treatment allocation introduce confounding and selection bias.
- Total cohort size and radiographic criteria standardization are not fully detailed.
Future Directions: Prospective, ideally randomized, trials to validate nitrogen washout efficacy and safety; standardized protocols for SP severity and imaging-based resolution.
BACKGROUND: Neonatal pneumothorax is dependent on severity of symptoms and leak size. Treatment of Pneumothorax with Nitrogen washout is controversial. OBJECTIVE: To compare outcomes of neonates treated for spontaneous pneumothorax (SP) compared with those managed expectantly. METHODS: A retrospective review of medical records of neonates (≥34weeks) diagnosed with SP born between 2011 and 2021. Neonatal characteristics and outcomes were compared between neonates treated for SP with either targeted oxygen therapy (to stabilize saturation ≥93%) or nitrogen washout and those under expectant management. RESULTS: Among neonates with respiratory distress and desaturation (n = 64), nitrogen washout reduced median time to resolution compared to targeted oxygen therapy (31 [12-67] vs 81 [31.8-123.5] hours, p = 0.012). Expectantly managed neonates with distress (n = 87) experienced delayed feeding initiation, while time to resolution and hospital stay were similar. CONCLUSION: Nitrogen washout is superior to targeted oxygen therapy in time to resolution in infants presenting with respiratory distress and desaturation.
3. Lung Ultrasound as a Tool for Analysis of Ventilation in Children With Respiratory Failure.
In a longitudinal PICU cohort (n=17) of invasively ventilated children, lung ultrasound using the BLUE protocol showed reduced A-lines, worsened B-lines, and significant decreases in pleural sliding and bat sign, tracking ventilatory course and potential complications. Findings support LUS as a bedside monitoring tool during hospitalization and weaning.
Impact: Demonstrates actionable LUS metrics linked to ventilation trajectory in critically ill children, informing ICU monitoring and weaning strategies.
Clinical Implications: Incorporate standardized LUS (BLUE protocol) to monitor pleural sliding and B-lines during invasive ventilation and weaning to detect complications early and guide ventilatory adjustments.
Key Findings
- In 17 ventilated children, A-lines decreased and B-lines worsened over time, indicating evolving lung pathology.
- Pleural sliding decreased from 14 to 3 (p=0.04) and bat sign from 10 to 5 (p=0.002), while stratospheric sign trended down (p=0.08).
- Moderate negative correlations were observed between MV duration and A-lines and bat sign, supporting LUS as a monitoring tool.
Methodological Strengths
- Prospective longitudinal assessment using a standardized ultrasound protocol (BLUE).
- Clinically relevant serial measurements during hospitalization and weaning.
Limitations
- Small single-center sample (n=17) limits generalizability and statistical power.
- Some reported correlations/sign trends are exploratory and require confirmation against gold-standard imaging.
Future Directions: Larger multicenter prospective studies integrating LUS with clinical outcomes and imaging to establish thresholds for ventilator adjustments and weaning readiness.
OBJECTIVE: To describe lung ultrasound findings in children with respiratory failure on invasive mechanical ventilation (MV). METHOD: This is a longitudinal, observational, quantitative study conducted in the pediatric intensive care unit. Children with respiratory distress syndrome, aged between 6 months and 7 years, on invasive MV were included in the study. Lung ultrasound was performed using the BLUE protocol in the first 48 h of hospitalization and during ventilatory weaning. RESULTS: Seventeen patients who presented a significant reduction in A lines were included in the study. B lines showed worsening, indicating possible pulmonary complications. The decrease in pleural sliding from 14 to 3 (p = 0.04) and in the bat sign from 10 to 5 (p = 0.002) was statistically significant. The stratospheric sign showed a favorable trend (reduction from 3 to 0), but the p value of 0.08 was not significant. There was a moderate negative correlation between MV time and A lines, while a moderate positive correlation was observed between MV time and A lines. Furthermore, a moderate negative correlation between MV time and bat sign was also significant. CONCLUSION: It is indicated that bedside lung ultrasound is a valuable tool for monitoring and management of children on invasive MV, allowing the follow-up of critical pediatric patients during the hospitalization period.