Daily Respiratory Research Analysis
Analyzed 140 papers and selected 3 impactful papers.
Summary
Three impactful respiratory studies emerged: an expert Delphi established consensus criteria for defining ARDS resolution, a PRISMA- and PROSPERO-registered meta-analysis clarified benefits and limits of LISA/MIST in extremely preterm infants, and a large reader study showed an AI tool (SOFIA) improves diagnostic agreement and prognostic accuracy for UIP on HRCT. Collectively, they standardize endpoints, refine neonatal respiratory care choices, and enhance imaging-based diagnosis in fibrotic ILD.
Research Themes
- Consensus endpoints in critical care (ARDS resolution definition)
- Noninvasive neonatal respiratory strategies (LISA/MIST) and outcomes
- AI-assisted imaging diagnostics for interstitial lung disease (UIP classification)
Selected Articles
1. Deep-Learning Algorithm Diagnostic Support for Usual Interstitial Pneumonia Pattern Recognition in Fibrotic Interstitial Lung Disease.
In a global reader study of 312 clinicians reviewing 203 HRCTs, providing SOFIA outputs improved inter-observer agreement for definite UIP and enhanced clinicians’ prognostic accuracy. The AI tool supported guideline-based classification, reducing variability in HRCT interpretation for fibrotic ILD.
Impact: Improved diagnostic consistency for UIP can standardize care pathways and trial eligibility, addressing a key bottleneck in ILD management and research.
Clinical Implications: AI-assisted HRCT review may be integrated as a decision support tool to harmonize UIP classification, potentially improving multidisciplinary diagnoses and patient selection for antifibrotic therapy and trials.
Key Findings
- Providing SOFIA algorithm outputs to reviewers improved inter-observer diagnostic agreement for definite UIP on HRCT.
- Clinicians’ prognostic accuracy improved after access to SOFIA outputs.
- The study involved 312 reviewers across 49 countries and 203 HRCTs from a national fibrotic ILD registry.
Methodological Strengths
- Large, international reader cohort (312 reviewers) with pre/post AI exposure
- Guideline-concordant AI tool evaluated against real-world HRCTs (n=203)
Limitations
- Non-randomized reader study; potential anchoring to AI outputs
- Generalizability depends on imaging acquisition quality and AI training data
Future Directions: Prospective workflow-integrated trials assessing impact on MDT decisions, treatment selection, and patient outcomes; external validation across diverse scanners and populations.
BACKGROUND AND OBJECTIVE: High resolution computed tomography (HRCT) scan diagnostic classification for usual interstitial pneumonia (UIP) plays a critical role in therapeutic decision-making and clinical trial eligibility for interstitial lung disease (ILD) patients, but is subject to variability. A deep learning algorithm, the Systematic Objective Fibrotic Imaging Analysis Algorithm (SOFIA), has been validated to assist classification of HRCTs based on current guidelines. In this study, we evaluate the impact of SOFIA on inter-observer agreem
2. Less invasive surfactant administration (LISA) in extreme preterm infants - A systematic review and meta-analysis.
Across 26 studies (12 meta-analyzed), LISA/MIST versus INSURE showed no difference in BPD or death but reduced mechanical ventilation within 72 hours; this MV effect was not seen when restricting to RCTs. Compared with intubation, LISA/MIST reduced BPD, MV, death, IVH, pneumothorax, and ROP, though findings require cautious interpretation in extremely preterm infants.
Impact: This synthesis clarifies where LISA/MIST confers advantages and where evidence remains limited, informing protocol selection in extremely preterm infants.
Clinical Implications: In centers with LISA/MIST capability, prioritizing non-intubation strategies can reduce early mechanical ventilation and several morbidities versus intubation, while recognizing no clear BPD/mortality benefit over INSURE in <28-week infants.
Key Findings
- LISA/MIST vs INSURE: no significant difference in BPD or death; reduced MV within 72 hours (RR 0.70, 95% CI 0.55–0.90), an effect not seen when analyzing only RCTs.
- LISA/MIST vs intubation: reduced BPD (RR 0.76), MV (RR 0.61), death (RR 0.63), IVH (RR 0.62), pneumothorax (RR 0.58), and ROP (RR 0.61).
- PRISMA-compliant, PROSPERO-registered (CRD42025630748) with GRADE assessment; no RCT directly comparing LISA/MIST with INSURE in <28-week infants.
Methodological Strengths
- PRISMA-guided, PROSPERO-registered meta-analysis with GRADE certainty assessment
- Broad comparative framework: LISA/MIST vs INSURE and vs intubation/MV
Limitations
- No direct RCT of LISA/MIST vs INSURE in infants <28 weeks’ gestation
- Potential heterogeneity and limited power; composite BPD/death analysis infeasible
Future Directions: Head-to-head RCTs of LISA/MIST vs INSURE in extremely preterm infants and standardized outcome sets (including composite BPD/death) to resolve competing risk dynamics.
Background Surfactant therapy is performed in different manners in respiratory distress syndrome. Surfactant therapy in spontaneously breathing infants <28weeks' gestation (LIST/MIST) was compared to intubation with different periods of mechanical ventilation (MV/intubation; INSURE) for efficacy and safety. Methods This systematic review and meta-analysis followed PRISMA guidelines with PROSPERO registration (CRD42025630748) using GRADE recommendations. Primary outcomes were BPD at 36weeks' postmenstrual age, MV in first 72hours; and deat
3. Defining the Resolution of Acute Respiratory Distress Syndrome: A Delphi Consensus Study.
A 19-expert modified Delphi reached high consensus on two core criteria for ARDS resolution: sustained resolution of hypoxemia (PaO2/FiO2 > 300 or SpO2/FiO2 > 315 for >24 h) and normalization of respiratory support (minimal ventilatory assistance if intubated, or no longer requiring CPAP/NIV/HFNC/oxygen depending on peak support). This provides a pragmatic, harmonized endpoint for research and clinical reporting.
Impact: A consensus definition addresses a major gap in ARDS trials and observational studies, enabling consistent endpoint adjudication and comparability across interventions.
Clinical Implications: Clinicians and investigators can adopt these criteria to define ARDS resolution in protocols, registries, and quality metrics, facilitating benchmarking and trial design.
Key Findings
- Consensus elements: hypoxemia resolution defined as PaO2/FiO2 > 300 (or SpO2/FiO2 > 315) sustained for >24 hours.
- Normalization of respiratory support: minimal ventilatory assistance if intubated; discontinuation of CPAP/NIV, HFNC, or oxygen according to prior maximal support.
- Three-round modified Delphi with 19 ARDS experts achieved predefined agreement thresholds.
Methodological Strengths
- Prespecified consensus thresholds and iterative refinement across three rounds
- Panel selected by predefined expertise criteria
Limitations
- Expert consensus without prospective validation against outcomes
- Potential panel and selection bias; limited number of experts (n=19)
Future Directions: Validate definition performance in cohorts and trials; assess prognostic discrimination and responsiveness to interventions; refine for pediatric and resource-limited settings.
OBJECTIVES: There are no established criteria to define the resolution of acute respiratory distress syndrome (ARDS). We aimed to develop an expert consensus definition of ARDS resolution. DESIGN: Modified Delphi consensus study with three iterative rounds. SETTING: Electronic surveys. SUBJECTS: A panel of 19 ARDS experts participated in the Delphi process. Experts were identified using prespecified criteria. INTERVENTIONS: The Delphi process was conducted over three rounds. Item generation was performed in round 1 with all panelists in