Lung-Protective Ventilation for Pediatric Acute Respiratory Distress Syndrome: A Nonrandomized Controlled Trial.

Abstract

Despite the recommendation for lung-protective mechanical ventilation (LPMV) in pediatric acute respiratory distress syndrome (PARDS), there is a lack of robust supporting data and variable adherence in clinical practice. This study evaluates the impact of an LPMV protocol vs. standard care and adherence to LPMV elements on mortality. We hypothesized that LPMV strategies deployed as a pragmatic protocol reduces mortality in PARDS. Multicenter prospective before-and-after comparison design study. Twenty-one PICUs. Patients fulfilled the Pediatric Acute Lung Injury Consensus Conference 2015 definition of PARDS and were on invasive mechanical ventilation. The LPMV protocol included a limit on peak inspiratory pressure (PIP), delta/driving pressure (DP), tidal volume, positive end-expiratory pressure (PEEP) to F io2 combinations of the low PEEP acute respiratory distress syndrome network table, permissive hypercarbia, and conservative oxygen targets. There were 285 of 693 (41·1%) and 408 of 693 (58·9%) patients treated with and without the LPMV protocol, respectively. Median age and oxygenation index was 1.5 years (0.4-5.3 yr) and 10.9 years (7.0-18.6 yr), respectively. There was no difference in 60-day mortality between LPMV and non-LPMV protocol groups (65/285 [22.8%] vs. 115/406 [28.3%]; p = 0.104). However, total adherence score did improve in the LPMV compared to non-LPMV group (57.1 [40.0-66.7] vs. 47.6 [31.0-58.3]; p < 0·001). After adjusting for confounders, adherence to LPMV strategies (adjusted hazard ratio, 0.98; 95% CI, 0.97-0.99; p = 0.004) but not the LPMV protocol itself was associated with a reduced risk of 60-day mortality. Adherence to PIP, DP, and PEEP/F io2 combinations were associated with reduced mortality. Adherence to LPMV elements over the first week of PARDS was associated with reduced mortality. Future work is needed to improve implementation of LPMV in order to improve adherence.