Seeing readiness: a pilot diagnostic accuracy study integrating diaphragm and lung ultrasound to predict extubation...

Predicting extubation readiness in the neonatal intensive care unit (NICU) remains difficult. Prolonged mechanical ventilation may contribute to ventilator-induced diaphragmatic dysfunction, whereas premature extubation increases morbidity. We evaluated whether diaphragm ultrasound (DUS) parameters, alone or integrated with lung ultrasound aeration scoring, predict extubation outcomes in ventilated neonates. In this prospective, single-center pilot diagnostic accuracy study, 72 mechanically ventilated neonates underwent a standardized spontaneous breathing trial (SBT) on a Hamilton-C1 Neo ventilator while intubated. Index tests included diaphragmatic excursion (DE), inspiratory and expiratory diaphragm thickness (DTi, DTe), and inspiratory velocity, alongside lung aeration assessment using the LUS aeration score (LUS, 0-18; 6 regions) and the extended LUS aeration score (eLUS, 0-30; 10 regions). The reference standard was extubation outcome, with extubation failure defined as reintubation within 48 h. Discrimination was assessed using receiver operating characteristic (ROC) analyses (AUC with 95% confidence intervals), with DeLong tests for pairwise AUC comparisons. Extubation succeeded in 53/72 (73.6%) and failed in 19/72 (26.4%). Success was associated with higher DE (7.82 ± 2.48 vs 4.08 ± 1.10 mm; p < 0.001) and DTi (2.19 ± 0.48 vs 1.16 ± 0.39 mm; p < 0.001), and with lower LUS (3.79 ± 1.79 vs 6.01 ± 1.20; p < 0.001) and eLUS (4.09 ± 2.10 vs 8.11 ± 2.64; p < 0.001). ROC analyses showed excellent discrimination for DE (AUC 0.978; 95% CI 0.912-0.998) and DTi (AUC 0.928; 95% CI 0.841-0.975), and good discrimination for inspiratory velocity (AUC 0.843; 95% CI 0.738-0.918). DE outperformed inspiratory velocity (ΔAUC 0.135; p = 0.001), while DE and DTi were similar (ΔAUC 0.0506; p = 0.179). For predicting failure, eLUS (AUC 0.876; 95% CI 0.778-0.942) and LUS (AUC 0.838; 95% CI 0.732-0.914) did not differ (ΔAUC 0.0387; p = 0.178). After adjustment for gestational age, higher LUS/eLUS and lower DTi remained independently associated with failure.

In this pilot study, DUS, particularly DE, provided excellent discrimination for extubation outcome, while LUS/eLUS added complementary information on residual lung disease burden. Multicenter validation is needed before routine clinical adoption.

• Prolonged mechanical ventilation may contribute to ventilator-induced diaphragmatic dysfunction; diaphragmatic performance is therefore relevant to weaning success. • Lung ultrasound aeration scoring can quantify residual lung disease burden and has been investigated as a predictor of post-extubation respiratory deterioration.

• In this single-center pilot diagnostic accuracy study, diaphragmatic excursion measured during a standardized SBT showed excellent discrimination for extubation success within 48 h. • Lung aeration scores provided complementary prediction for extubation failure, with similar performance between eLUS and LUS.