Daily monitoring of lung aeration during pediatric acute respiratory syndrome (PARDS) is a common practice in pediatric intensive care unit and several monitoring tools have been used at bedside.1 Recently, lung ultrasound imaging has gained popularity as a diagnostic and monitoring tool of lung aeration in neonates and children offering fast, radiation-free, and real-time images at bedside.2,3 A semiquantitative lung ultrasound score (LUS) has been developed to monitor lung aeration in adult patients with ARD4,5 and has been used also during extracorporeal membrane oxygenation (ECMO).6,7 However, as no data exist about the use of LUS during pediatric ECMO, we aimed to evaluate its feasibility and potential to assess lung aeration in 20 patients with PARDS supported with ECMO. In this retrospective observational cohort, LUS was calculated daily from the ECMO beginning till weaning or withdrawal. LUS was computed by one consultant (Dr Di Nardo) for all the patients included in the study using a transverse scan with a 2.5 MHz convex probe. Right and left hemithorax were divided in three areas for each side by the parasternal, anterior axillary, posterior axillary, and paravertebral lines. Each areas was subsequently divided in upper and lower zones (Figure 1A).6,7 A total of 12 lung zones were examined. Each zone was scored according to the worst lung ultrasound findings detected. Lung ultrasound findings were reported as follows (Figure 1B): score 0, normal lung aeration: presence of lung sliding (A lines or less than two isolated B lines); score 1, moderate loss of lung aeration: presence of multiple and spaced B1 lines developing from the pleural line or from small juxta-pleural consolidations; score 2, severe loss of lung aeration: presence of multiple and coalescent B2 lines developing from the pleural line or from small juxta-pleural consolidations; score 3, lung consolidation: presence of consolidated areas, including images representative of dynamic or static air bronchograms.6,7 LUS (range 0–36) was computed as the sum of the scores assigned to each zone.Figure 1.: (A) Parasternal, anterior axillary, posterior axillary, and paravertebral lines delineating three upper and three lower lung areas in each hemithorax. (B) LUS defined as normal aeration (score 0), moderate loss of lung aeration (score 1), severe loss of lung aeration (score 2), and lung consolidation (score 3). *B lines; C, consolidation; LUS, lung ultrasound score.All the patients were in supine position. Informed consent for data publication was obtained by parents or guardians. Quantitative variables were expressed as median and interquartile range (IQR) and analyzed with Mann-Whitney U test. Categorical variables were expressed as number and percentage and analyzed with χ2 or Fisher test. Changes overtime of the median LUS within survivors and nonsurvivors were analyzed with the Friedman test. Correlation between the daily median LUS and the median oxygenation index (OI) (calculated by considering the FiO2 averaged between the one delivered to the ECMO pump and to the ventilator) in survivors and nonsurvivors was analyzed using the Spearman analysis. P ≤ 0.05 were considered significant. All analyses were done using PRISM 8 (GraphPAD Software, San Diego, CA). Baseline patients’ characteristics, diagnosis, and the respiratory parameters at ECMO cannulation were reported in Table, Supplemental Digital Content 1, https://links.lww.com/ASAIO/A532, and Figures, Supplemental Digital Contents 1 and 2, https://links.lww.com/ASAIO/A532. Ten patients (50%) were supported with venovenous and the remaining ones with venoarterial ECMO. Fourteen patients (70%) were weaned from ECMO and survived to hospital discharge, while six (30%) died during ECMO. Baseline patients’ characteristics and the respiratory parameters at ECMO cannulation did not significantly differ between survivors and nonsurvivors. Figure 2 shows LUS in survivors and nonsurvivors from ECMO deployment to the 14th day. LUS at ECMO beginning (27.00 [IQR, 25.75–28.25] vs. 27.50 [IQR, 25.75–28.50]; p = 0.08) was not different between survivors and nonsurvivors. LUS reached its peak (30.00 [IQR, 29.00–32.25] and 33.00 [IQR, 31.00–34.00] in survivors and nonsurvivors respectively) after 3 days (IQR, 2–3) and was not different between survivors and nonsurvivors (p = 0.07). LUS when survivors were weaned from ECMO was 12.50 (IQR, 11.75–13.00).Figure 2.: Median LUS overtime in survivors and nonsurvivors. Symbols indicates p values between patients comparisons (survivors vs. nonsurvivors): # p = 0.04, Ø p = 0.02, ¤ p = 0.001. LUS, lung ultrasound score.LUS significantly decreased overtime only within survivors (p < 0.001), while remained stable within nonsurvivors (p = 0.114). Between survivors and nonsurvivors comparisons, LUS showed to be different from the fifth day of ECMO onward (p = 0.04). A significant negative correlation (r = –0.57; p = 0.02) between the daily LUS and the OI was shown only in survivors. Consistent with previous adult experiences,6,7 LUS seems a feasible tool to monitor lung aeration during pediatric ECMO. LUS significantly reduces in patients where lung aeration improves, while remains stable and high in patients that are not weaned from ECMO and subsequently died. Improvement of LUS overtime in survivors negatively correlates with an increase of the OI, which is common during the ECMO weaning phase and corresponds to an increase of the mechanical ventilation settings. In nonsurvivors, instead, both LUS and OI, remain stable overtime confirming no improvement of lung aeration and stable gas exchange only provided by ECMO. Since the retrospective nature of the study and the small number of patients included, any definitive clinical conclusion about the use of lung ultrasound during pediatric ECMO is precluded; however, these data support the feasibility of LUS and show the importance to monitor lung aeration during pediatric ECMO together with gas exchange especially in the weaning phase. These results may prompt future studies to investigate the use of LUS as a predictive tool for ECMO weaning in severe PARDS supported with ECMO.