Comparative Effectiveness of Protective Ventilation Strategies for Moderate and Severe ARDS: Network Meta-Analysis.

Abstract

RATIONALE\nChoosing the best ventilation strategy for acute respiratory distress syndrome (ARDS) is complex, yet highly relevant to clinicians during a respiratory pandemic.\n\n\nOBJECTIVES\nTo compare the effects of low tidal volumes (Vt), high Vt, high positive end-expiratory pressure (high PEEP), prone ventilation, high frequency oscillation (HFO), and venovenous extracorporeal membrane oxygenation (VV ECMO) on mortality in ARDS.\n\n\nMETHODS\nWe performed a network meta-analysis of randomized trials. We applied GRADE methodology to discern the relative effect of interventions on mortality.\n\n\nMEASUREMENTS AND MAIN RESULTS\nWe analyzed 34 trials including 9085 adults with primarily moderate to severe ARDS (median baseline PaO2/FiO2 118, interquartile range 110-143). Prone positioning combined with low Vt was the best strategy (relative risk [RR] 0.74, 95% CI 0.60-0.92 vs. low Vt; high certainty). VV ECMO also rated among the best (RR 0.78, 95% CI 0.58-1.05 vs. low Vt and 0.66, 95% CI 0.49-0.88 vs. high Vt), but with lower certainty since VV ECMO was restricted to very severe ARDS (mean baseline PaO2/FiO2≤75). High PEEP combined with low Vt rated intermediately (RR 0.91, 95% CI 0.81-1.03 vs. low Vt; low certainty; and RR 0.77, 95% CI 0.65-0.91 vs. high Vt; moderate certainty). High Vt rated worst (RR 1.19, 95% CI 1.02-1.37 vs. low Vt, moderate certainty), and we found no support for HFO or high Vt with prone ventilation.\n\n\nCONCLUSIONS\nThese findings suggest that combining low Vt with prone ventilation is associated with the greatest reduction in mortality for critically ill adults with moderate to severe ARDS.