Diaphragmatic ultrasound and thoracic fluid content for prediction of non-invasive ventilation failure in neonates: a randomized controlled trial.

Summary

In a randomized neonatal cohort (n=90), diaphragmatic thickening fraction and excursion measured within 3 hours of life accurately predicted NIV failure, outperforming thoracic fluid content, with a combined DTF+TFC model reaching AUC 0.93. Findings support early bedside ultrasound to stratify risk during initial NIV.

Key Findings

• DTF and DE were significantly higher in NIV successes than failures (all p<0.001).
• DTF and DE achieved AUCs of 0.90 and 0.89, outperforming TFC (AUC 0.81).
• DTF (adjusted OR 0.89, 95% CI 0.83–0.96) and DE (adjusted OR 0.38, 95% CI 0.20–0.73) independently predicted NIV failure.
• Combined DTF+TFC model provided the highest discrimination (AUC 0.93, 95% CI 0.89–0.97).

Clinical Implications

Incorporating early diaphragmatic ultrasound into standard neonatal NIV assessment may guide escalation decisions, optimize monitoring, and allocate higher-acuity care to high-risk infants.

Limitations

• Single-center, small sample size; limited power for clinical outcomes
• Trial registered late (August 22, 2025), raising concerns about preregistration and potential bias; external validation lacking

Future Directions

Multicenter validation, integration into decision algorithms for NIV escalation, and assessment of impact on clinical outcomes (intubation, BPD, mortality).