Abstract Rationale Acute respiratory distress syndrome (ARDS) affects a quarter million people in the United States annually, with 35-45% mortality and few effective treatments. While an initial landmark ARDS randomized controlled trial (RCT) showed mortality benefit with neuromuscular blockade, the subsequent Revaluation of Systemic Early Neuromuscular Blockade (ROSE) trial found no benefit. Thus, which patients benefit from neuromuscular blockade remains unclear, and is reflected in conflicting recommendations from leading society guidelines. Given this clinical equipoise, we hypothesize there is considerable variability in neuromuscular blockade use across hospitals. Characterizing current practices in a large, multi-center, real-world cohort is an essential first step in determining which patients benefit from neuromuscular blockade. Methods This retrospective observational study used electronic health record data from adult patients hospitalized 1/1/2018-12/31/2024 at 34 hospitals from seven health systems in the Common Longitudinal ICU Format consortium. We included patients on mechanical ventilation with moderate-to-severe hypoxemic respiratory failure within 5 days of intubation, based on ROSE RCT criteria: PaO2/FiO2 ratio (or equivalent SpO2/FiO2) ≤150 and PEEP≥8 cmH2O. The primary outcome was use of neuromuscular blockade within 48 hours of meeting eligibility, defined as continuous infusion or 2 bolus doses post-intubation. We analyzed medication use across individual hospitals and used random-effects meta-analysis to pool health system results in a multivariable logistic regression model that evaluated the association between pre- and post-ROSE trial publication (5/2019) and neuromuscular blockade use, adjusting for pre-specified covariates including COVID-19 infection. Results Among 27,176 patients with moderate-to-severe hypoxemia, the mean age was 60.4 years, with 39% female, 57% White, and 52% PaO2/FiO2100. Among these patients, 14.5% received neuromuscular blockade within 48 hours; 12.4% on continuous infusions and 4.6% with intermittent use. Use varied considerably across hospitals (median: 14.8%, IQR: 9.5-17.7%), in both academic (median: 14.9%, IQR: 10.8-17.7%) and community hospitals (median: 14.8%, IQR: 8.7-17.4%), and in groups categorized by total hospital ICU capacity. There was no difference in neuromuscular blockade use pre-ROSE (12.5%) versus post-ROSE (14.9%) either consortium-wide (adjusted OR: 1.02, 95% CI: 0.86-1.21) or within any individual health system. Conclusions In a multi-center description of 25,000 mechanically ventilated patients with moderate-to-severe hypoxemia, hospital-level neuromuscular blockade use varied widely, with no change after the publication of the negative ROSE RCT. These findings represent real-world implications of the current equipoise regarding optimal neuromuscular blockade use. Additional work to identify which patients these medications benefit is urgently needed to provide clear evidence-based recommendations in severe respiratory failure and ARDS. This abstract is funded by: NHLBI
Barker et al. (Fri,) studied this question.