Publication

• Title: Early Neuromuscular Blockade in the Acute Respiratory Distress Syndrome
• Acronym: ROSE (Reevaluation of Systemic Early Neuromuscular Blockade)
• Year: 2019
• Journal published in: New England Journal of Medicine
• Citation: Moss M, Huang DT, Brower RG, Ferguson ND, Ginde AA, Gong MN, et al; National Heart, Lung, and Blood Institute PETAL Clinical Trials Network. Early neuromuscular blockade in the acute respiratory distress syndrome. N Engl J Med. 2019;380:1997-2008.

Context & Rationale

• Background

  Neuromuscular blockade (NMBA) can reduce patient–ventilator dyssynchrony, facilitate lung-protective ventilation, and may mitigate ventilator-induced lung injury in early, severe ARDS.
  Prior randomised trials (notably ACURASYS) and smaller mechanistic studies suggested improved oxygenation and a potential mortality signal, but were conducted in a different sedation/ventilation era (routine deep sedation; variable PEEP strategies).
  Key trade-offs include the need for deep sedation (with attendant haemodynamic and delirium risks), immobility, and concern for ICU-acquired weakness.
  ROSE was explicitly designed as a contemporary “re-evaluation” of early continuous NMBA infusion within a protocolised lung-protective/high-PEEP strategy and a modern light-sedation comparator bundle.¹
• Research Question/Hypothesis

  In adults with moderate-to-severe ARDS early in the course of invasive ventilation, does early, continuous cisatracurium infusion (with mandated deep sedation) reduce 90-day in-hospital mortality compared with a usual-care strategy that avoids routine paralysis and targets lighter sedation?
• Why This Matters

  NMBA infusion is a high-leverage ICU intervention: if beneficial, it should be reliably implemented early; if neutral or harmful, routine use should be avoided.
  The trial directly tests whether previously observed benefits are robust to contemporary sedation practices and protocolised lung-protective/high-PEEP ventilation.
  Results have immediate implications for guidelines that previously supported early continuous paralysis in moderate-to-severe ARDS.

Design & Methods

• Research Question: Whether early, continuous cisatracurium infusion (48 hours) improves 90-day in-hospital survival in moderate-to-severe ARDS when compared with a usual-care strategy without routine paralysis and with lighter sedation targets.
• Study Type: Multicentre, randomised, parallel-group trial conducted in ICUs at 48 hospitals in the United States (PETAL network); open-label (no blinding of clinicians/participants); site-stratified randomisation; investigator-initiated.
• Population:
  • Setting: ICU patients receiving invasive mechanical ventilation.
  • Key inclusion: Moderate-to-severe ARDS early after onset (enrolment within 48 hours of meeting ARDS criteria); PaO₂:FiO₂ ≤150 with PEEP ≥8 cm H₂O; bilateral infiltrates consistent with ARDS.
  • Key exclusions: Not fully reported in the published manuscript summary here; protocolised exclusions included conditions where deep sedation/paralysis or study procedures would be unsafe or confounded (e.g., baseline neuromuscular disease, pregnancy, contraindications to cisatracurium or deep sedation).
• Intervention:
  • NMBA strategy: Initiation targeted within hours of randomisation; cisatracurium administered as a bolus followed by continuous infusion for 48 hours (protocolised dosing; escalation permitted for inadequate paralysis).
  • Sedation bundle: Deep sedation mandated during paralysis (Richmond Agitation–Sedation Scale targets in the deep range).
  • Ventilation bundle: Protocolised lung-protective ventilation with low tidal volume and high-PEEP strategy.
• Comparison:
  • Usual-care strategy: Avoid routine continuous NMBA infusion; target lighter sedation (with structured sedation guidance).
  • Rescue NMBA: Clinicians could use NMBA when clinically indicated (e.g., severe dyssynchrony or refractory gas exchange), leading to partial crossover exposure.
  • Ventilation: Same protocolised lung-protective/high-PEEP approach.
• Blinding: Unblinded intervention delivery; primary outcome (mortality) is objective, but co-interventions (sedation intensity, mobilisation practices) were susceptible to performance bias.
• Statistics: A total sample size of 1,408 patients was planned to detect an 8% absolute reduction in 90-day in-hospital mortality (from 35% to 27%) with 90% power at a two-sided 5% significance level; primary analysis was intention-to-treat with a group-sequential monitoring plan.
• Follow-Up Period: In-hospital outcomes through day 90; additional survivorship and functional follow-up out to 12 months in supplementary reporting.²

Key Results

This trial was stopped early. Recruitment was halted after 1,006 patients at the second interim analysis because the Data and Safety Monitoring Board recommended stopping for futility (no emerging signal of benefit on the primary endpoint).

• Routine early continuous cisatracurium infusion (with mandated deep sedation) did not improve 90-day in-hospital mortality compared with a strategy avoiding routine paralysis (42.5% vs 42.8%; difference -0.3 percentage points; 95% CI -6.4 to 5.9; P=0.93).
• Physiological/organ-support endpoints were closely overlapping (e.g., ventilator-free days through day 28: 8.8 ± 10.6 vs 8.8 ± 10.8; difference -0.1; 95% CI -1.4 to 1.2; P=0.97).
• Safety signals favoured the usual-care strategy: serious cardiovascular adverse events were more frequent with early NMBA (14 vs 4 events; P=0.02), and early vasopressor escalation was more common in supplementary reporting (45.9% vs 36.6%; difference 9.3 percentage points; 95% CI 3.2 to 15.3).²

Internal Validity

• Randomisation and allocation: Central, site-stratified randomisation supports good allocation concealment; baseline characteristics were broadly balanced, though minor imbalances were present (e.g., small differences in shock prevalence and ARDS-to-randomisation interval were reported).
• Early stopping: Planned sample size was 1,408; stopping at 1,006 reduces precision for modest benefits/harms and increases the chance of over-interpreting secondary outcomes.
• Performance/detection bias: Open-label delivery creates a material risk that co-interventions differed (sedation depth, mobilisation, ventilator management decisions), particularly for subjective/behavioural endpoints.
• Protocol adherence: Cisatracurium infusion was started in 488/501 (97.4%) in the intervention group; median time to start infusion was 1.9 ± 1.4 hours; 336/451 (74.5%) completed the 48-hour infusion, with common early cessation due to protocolised stop-allowance criteria or death.²
• Separation of the variable of interest: Clear sedation separation early, but incomplete isolation of “paralysis” from “sedation strategy”.
  
  • Sedation depth: Day 1 mean RASS -4.8 ± 0.4 vs -2.7 ± 1.6 (difference -2.1; 95% CI -2.2 to -2.0); Day 2 mean RASS -4.8 ± 0.5 vs -2.5 ± 1.5 (difference -2.3; 95% CI -2.4 to -2.2).²
  • Crossover NMBA exposure: Usual-care patients received non-cisatracurium NMBA during the first 48 hours in 86/505 (17.0%), indicating partial treatment contamination; additional NMBA exposure after the first 48 hours was also documented in both groups.²
• Timing: The intervention was delivered early by design (rapid initiation after randomisation), aligned with the hypothesised “early lung injury” window.
• Dose: Fixed-dose cisatracurium with escalation was used; actual delivered dose and duration were substantial (e.g., total cisatracurium over the first 48 hours ≈1,807 mg among those with recorded totals), but a sizeable proportion stopped early per protocol/clinical course.²
• Adjunctive therapy use: Key ARDS adjuncts (e.g., prone positioning, inhaled vasodilators, ECMO, glucocorticoids) were reported and were broadly similar between groups in supplementary tables, reducing (but not eliminating) confounding from differential rescue strategies.²
• Outcome assessment: Mortality is objective; long-term patient-reported outcomes and some functional assessments were collected with structured follow-up, but ICU-level physical activity measures are susceptible to unblinded care-team behaviour.
• Statistical rigour: Intention-to-treat analysis was appropriate; prespecified subgroup analyses were performed without multiplicity adjustment (interpretation should remain cautious).

Conclusion on Internal Validity: Overall, internal validity appears moderate: randomisation and objective primary outcome support credibility, but open-label delivery, partial crossover, bundled sedation differences, and early stopping constrain causal attribution and the ability to detect small-but-real effects.

External Validity

• Population representativeness: Enrolled patients had moderate-to-severe ARDS early after onset and were managed in US hospitals with protocolised lung-protective/high-PEEP ventilation; this resembles many high-resource ICUs, but excludes patients outside early ARDS windows and those in settings without protocolised ventilation/sedation infrastructure.
• Intervention feasibility: Continuous cisatracurium infusion with mandated deep sedation is feasible in most ICUs, but practice patterns (drug availability, sedation culture, mobilisation staffing, haemodynamic monitoring) vary widely.
• Applicability across systems: The “usual-care” comparator explicitly targeted lighter sedation and tolerated rescue NMBA; applicability is strongest where light-sedation strategies are achievable and where alternative ARDS adjuncts (proning, ECMO) are available.
• Subpopulations: Findings apply most directly to early, moderate-to-severe ARDS; extrapolation to mild ARDS, very late ARDS, non-invasive support, or profoundly unstable patients should be cautious.

Conclusion on External Validity: External validity is good for contemporary, high-resource ICUs managing early moderate-to-severe ARDS with protocolised ventilation and sedation, but may be limited in settings where sedation/mobilisation practices or rescue adjunct availability differ substantially.

Strengths & Limitations

• Strengths:
  • Large, multicentre randomised trial in a contemporary ARDS care context (protocolised low tidal volume + high PEEP).
  • Clinically meaningful primary endpoint (90-day in-hospital mortality) with objective ascertainment.
  • Rich supplementary reporting, including sedation separation metrics, co-interventions, haemodynamic effects, and longer-term outcomes.²
• Limitations:
  • Open-label design with a bundled intervention (NMBA + mandated deep sedation) versus a bundled comparator (avoid routine NMBA + lighter sedation), limiting isolation of the “pure” NMBA effect.
  • Stopped early for futility, reducing power to detect modest effects and amplifying uncertainty around secondary endpoints.
  • Partial crossover NMBA exposure in the control arm (clinician-directed rescue), potentially diluting treatment separation.
  • Functional/rehabilitation outcomes are sensitive to unblinded practice patterns and have incomplete assessment in the sickest patients (death/early discharge).

Interpretation & Why It Matters

• Clinical practice

  In early moderate-to-severe ARDS managed with contemporary lung-protective/high-PEEP ventilation, routine early continuous cisatracurium infusion does not confer a survival advantage and should not be used as default therapy; NMBA can be reserved for selected rescue indications (severe dyssynchrony, refractory gas exchange) within a broader ARDS bundle.
• Mechanism vs bundle

  The trial tests a clinically realistic package (paralysis + deep sedation) against a competing package (avoid routine paralysis + lighter sedation), highlighting that any theoretical lung-protective benefit of paralysis may be offset by deep-sedation/immobility/haemodynamic costs.
• Safety signal

  Increased early vasopressor escalation and serious cardiovascular adverse events with early NMBA underscore the need for careful haemodynamic surveillance and reinforce the value of prioritising alternatives (optimised synchrony strategies, prone positioning) when feasible.²

Controversies & Subsequent Evidence

• “NMBA effect” vs “sedation strategy effect”: ROSE intentionally contrasted deep-sedation paralysis against a light-sedation, no-routine-paralysis strategy; this strengthens pragmatic relevance but limits mechanistic attribution, and was a focal point of correspondence critique.³
• Stopping for futility: The protocolised monitoring plan did not include a formal futility boundary, yet the DSMB recommended stopping for futility at the second interim analysis; this choice reduces the likelihood of detecting smaller but clinically important benefits (or delayed harms) and remains a recurrent methodological discussion point in post-publication commentary.³
• Subgroup signals: Prespecified subgroup analyses were largely neutral; a statistical interaction by Hispanic ethnicity was reported in supplementary subgroup tables (interaction P=0.015) without multiplicity control, and should be interpreted as hypothesis-generating rather than practice-changing.²
• Meta-analytic synthesis after ROSE: Multiple post-ROSE meta-analyses incorporating ROSE and earlier NMBA trials consistently show little-to-no mortality benefit with routine early continuous NMBA infusion, with some analyses suggesting reduced barotrauma but uncertain net clinical benefit.⁴⁵
• Guideline evolution: Contemporary guidelines and rapid practice recommendations published after ROSE generally move away from routine early continuous NMBA infusion in ARDS, favouring selective use (often intermittent boluses) in patients with severe dyssynchrony or refractory hypoxaemia while prioritising lung-protective ventilation and prone positioning.⁶⁷⁸

Summary

• ROSE randomised 1,006 patients with early moderate-to-severe ARDS to 48-hour cisatracurium infusion + deep sedation vs usual care avoiding routine paralysis with lighter sedation targets.
• The trial was stopped early for futility and showed no difference in 90-day in-hospital mortality (42.5% vs 42.8%; difference -0.3 percentage points; 95% CI -6.4 to 5.9; P=0.93).
• Key organ-support outcomes were similar (ventilator-free days through day 28: 8.8 ± 10.6 vs 8.8 ± 10.8; difference -0.1; 95% CI -1.4 to 1.2; P=0.97).
• Early NMBA strategy was associated with a signal of harm: more serious cardiovascular adverse events (14 vs 4 events; P=0.02) and more early vasopressor escalation in supplementary reporting.
• ROSE shifted practice and guideline recommendations away from routine early continuous NMBA infusion, positioning NMBA as a selective rescue therapy within a modern ARDS bundle.

Further Reading

Other Trials

• 2004Gainnier M, Roch A, Forel JM, et al. Effect of neuromuscular blocking agents on gas exchange in patients presenting with acute respiratory distress syndrome. Crit Care Med.
• 2006Forel JM, Roch A, Marin V, et al. Neuromuscular blocking agents decrease inflammatory response in patients presenting with acute respiratory distress syndrome. Crit Care Med.
• 2010Papazian L, Forel JM, Gacouin A, et al. Neuromuscular blockers in early acute respiratory distress syndrome. N Engl J Med. 2010;363:1107-1116.
• 2000The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342:1301-1308.
• 2013Guérin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013;368:2159-2168.

Systematic Review & Meta Analysis

• 2020Tarazan N, et al. Neuromuscular blocking agents in acute respiratory distress syndrome: a systematic review and meta-analysis. Crit Care.
• 2021Torbic H, et al. Neuromuscular blocking agents for acute respiratory distress syndrome: a systematic review and meta-analysis. Respir Care.
• 2020Shao C, et al. Neuromuscular blocking agents in ARDS: systematic review and meta-analysis. BMJ Open.
• 2023Savoie-White K, et al. Neuromuscular blockade and outcomes in ARDS: updated systematic review and meta-analysis. Heart & Lung.
• 2019Wei Q, et al. Neuromuscular blockade in ARDS: systematic review and meta-analysis. Front Pharmacol.

Observational Studies

• 2016Bellani G, Laffey JG, Pham T, et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries. JAMA. 2016;315:788-800.
• 2020Duggal A, Rezoagli E, Pham T, et al. Patterns of use of adjunctive therapies in patients with early moderate to severe acute respiratory distress syndrome. Chest. 2020;157:1497-1505.
• 2021Admon AJ, et al. Variation in early management practices in moderate-to-severe ARDS in the United States: the Severe ARDS: Generating Evidence Study. Chest. 2021;160:1304-1315.
• 2018Sottile PD, et al. Lower driving pressure and neuromuscular blocker use are associated with lower mortality in ARDS. Am J Respir Crit Care Med.
• 2022Lee Y, et al. Driving pressure, neuromuscular blockade, and mortality in patients with ARDS. Respir Care.

Guidelines

• 2020Alhazzani W, et al. Neuromuscular blockade in patients with ARDS: a rapid practice guideline. Intensive Care Med.
• 2021Evans L, Rhodes A, Alhazzani W, et al. Surviving Sepsis Campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med.
• 2023Cammarota G, et al. Management of acute respiratory distress syndrome: clinical practice guidelines by the European Society of Intensive Care Medicine. Intensive Care Med.
• 2024Qadir N, et al. An update on ARDS clinical practice guidelines. Am J Respir Crit Care Med.
• 2022Fujishima S, et al. Japanese Clinical Practice Guideline for management of acute respiratory distress syndrome: 2022 update. Respir Investig.

Notes

• Effect measures are presented as reported by the trialists (predominantly risk differences and mean differences); no additional calculations were performed.

Overall Takeaway

ROSE is a landmark “practice-resetting” trial: in a modern ARDS bundle that targets lighter sedation and uses protocolised lung-protective/high-PEEP ventilation, routine early continuous cisatracurium infusion with deep sedation did not improve survival and introduced plausible safety and recovery trade-offs. It repositioned neuromuscular blockade from a routine early strategy to a selective rescue tool integrated with other high-value ARDS therapies.

Bibliography

• 1.Huang DT, Angus DC, Thompson BT, et al. Design and rationale of the Reevaluation of Systemic Early Neuromuscular Blockade trial for acute respiratory distress syndrome. Ann Am Thorac Soc. 2017.
• 2.Moss M, Huang DT, Brower RG, et al. Early neuromuscular blockade in the acute respiratory distress syndrome. Supplementary Appendix. N Engl J Med. 2019.
• 3.Brink M, de Moraes AG, El-Yafawi I, Oeckler RA. Early neuromuscular blockade in the acute respiratory distress syndrome. N Engl J Med. 2019.
• 4.Tarazan N, et al. Neuromuscular blocking agents in acute respiratory distress syndrome: a systematic review and meta-analysis. Crit Care. 2020.
• 5.Torbic H, et al. Neuromuscular blocking agents for acute respiratory distress syndrome: a systematic review and meta-analysis. Respir Care. 2021.
• 6.Alhazzani W, et al. Neuromuscular blockade in patients with ARDS: a rapid practice guideline. Intensive Care Med. 2020.
• 7.Cammarota G, et al. Management of acute respiratory distress syndrome: clinical practice guidelines by the European Society of Intensive Care Medicine. Intensive Care Med. 2023.
• 8.Qadir N, et al. An update on ARDS clinical practice guidelines. Am J Respir Crit Care Med. 2024.