Effect of therapeutic interventions combined with lung-protective ventilation on long-term mortality of patients with acute respiratory distress syndrome: a network meta-analysis.

PURPOSE: Long-term effects of ventilatory strategies/adjunctive therapies received in the intensive care unit on mortality of patients with acute respiratory distress syndrome (ARDS) is uncertain. To explore 180-day mortality in adult patients with ARDS, we conducted a network meta-analysis (NMA) comparing the effects of various prespecified interventions added to lung-protective ventilation (LPV).

SOURCE: We systematically searched six databases on 8 November 2024. Two reviewers independently identified eligible randomized clinical trials with published Kaplan-Meier curves, exploring prespecified interventions combined with LPV and LPV alone. Data were synthesized with NMA of survival curves using Bayesian random effects fractional polynomial models.

PRINCIPAL FINDINGS: Twenty-two trials with 8,653 participants assessed six different interventions added to LPV (open lung strategy, neuromuscular blockade [NMBA], corticosteroids, high-frequency oscillatory ventilation [HFOV], prone positioning, and venovenous extracorporeal membrane oxygenation [VV ECMO]), compared with LPV alone. We did not include inhaled pulmonary vasodilator trials. The primary NMA computed survival probability of each intervention on the basis of survival probability of LPV alone at 180 days (0.52; 95% credible interval, 0.49 to 0.55). Compared with LPV alone, 1) the evidence suggests prone positioning results in a reduction in 180-day mortality; 2) an open lung strategy does not reduce 180-day mortality; 3) VV ECMO, NMBA, and corticosteroids may reduce 180-day mortality (very uncertain); and 4) HFOV may increase 180-day mortality (very uncertain). We extrapolated reported mortalities to 180-day mortality in 16 trials, where there might be discrepancy between raw and extrapolated numbers in event rate.

CONCLUSIONS

Prone positioning may improve long-term mortality in patients with ARDS.

STUDY REGISTRATION: PROSPERO ( CRD42019131849 ); first submitted 26 April 2019.