ABSTRACT
OBJECTIVE: To describe lung mechanics in Pediatric Acute Respiratory Distress Syndrome (PARDS) associated with acute COVID-19 and MIS-C with respiratory failure. METHODS: A concurrent multicenter observational study was performed, analyzing clinical variables and pulmonary mechanics of PARDS associated with COVID-19 in 4 Pediatric intensive care units (PICU) in Peru. The subgroup analysis included PARDS associated with multisystem inflammatory syndrome in children (MIS-C), MIS-PARDS, and PARDS with COVID-19 primary respiratory infection, C-PARDS. In addition, receiver operating characteristic (ROC) curve analysis for mortality and lung mechanics was performed. RESULTS: 30 patients were included. The age was 7.5 (4-11) years, 60% were male, and mortality was 23%. 47% corresponded to MIS-PARDS and 53% to C-PARDS groups. C-PARDS had positive RT-PCR in 67% and MIS-PARDS none (p < 0.001). C-PARDS group had more profound hypoxemia (P/F ratio < 100, 86% vs. 38%, p < 0.01) and higher driving-pressure [14(10-22) vs 10(10-12) cmH2O], and lower compliance of the respiratory system (CRS) [0.5 (0.3-0.6) vs 0.7(0.6-0.8) ml/ kg/cmH2O] compared with MIS-PARDS (all p < 0.05). The ROC analysis for mortality showed that driving pressure had the best performance [AUC 0.91(95%CI0.81-1.00), with the best cut-off point of 15 cmH2O (100% sensitivity and 87% specificity). Mortality in C-PARDS was 38% and 7% in MIS-PARDS (p = 0.09). MV-free days were 12(0-23) in C-PARDS and 23(21-25) in MIS-PARDS (p = 0.02). CONCLUSION: Patients with C-PARDS have lung mechanics characteristics similar to classic moderate to severe PARDS. This was not observed in patients with MIS-C. As seen in other studies, a driving pressure ≥ 15 cmH2O was the best discriminator for mortality. These findings may help guide ventilatory management strategies for these two different presentations.
