Editor's Note 

Inhaled Sedation Strategy for Acute Hypoxemic Respiratory Failure: New Evidence for Debate

Natalie L. McEvoy, Christopher W. Seymour

JAMA Published Online: October 14, 2026

doi: 10.1001/jama.2026.21349

Acute hypoxemic respiratory failure is common in the intensive care unit (ICU), and patients typically require sedation for ventilator-associated discomfort. Sedation using intravenous medications, particularly at high doses, is associated with important adverse effects, including hypotension, prolonged duration of mechanical ventilation, and delirium.¹ A potential alternative approach is to use inhaled anesthetic agents because they may have fewer adverse effects, but these agents are typically only used in the operating room, not in the ICU. In this issue of JAMA, Jerath and colleagues² report the findings of a pragmatic, open-label, randomized clinical trial (the SAVE-ICU study) to determine the effectiveness of a strategy of inhaled anesthesia plus intravenous sedation vs intravenous sedation alone. SAVE-ICU found that in mechanically ventilated adults with acute hypoxemic respiratory failure, inhaled sedation using isoflurane with intravenous adjuncts as clinically needed did not reduce hospital mortality compared with intravenous sedation (42.5% vs 43.8%; risk ratio, 0.98 [95% CI, 0.84 to 1.14]; P = .85). This Editor’s Note discusses several aspects of the study that factored into JAMA’s decision to publish this study.

First, this trial extends prior safety and efficacy data by using a pragmatic approach to examine the effectiveness of a novel sedation/anesthesia strategy. The previous SESAR trial explored the efficacy of sevoflurane vs propofol in patients with acute respiratory distress syndrome.³ Now the SAVE-ICU study evaluates whether these findings translate into improved patient outcomes in routine clinical practice. SAVE-ICU incorporates key pragmatic features, including flexible adherence, sedation targets, and delivery of the intervention in the ICU.⁴ Of course, pragmatic trials offer the potential for greater generalizability to clinical practice, but the design may include compromises in the fidelity with which interventions are delivered or monitored, potentially compromising interpretability. One example is the depth of sedation achieved in each group and the reporting of daily Richmond Agitation-Sedation Scale measurements. These bedside sedation targets were recorded in a limited fashion to minimize the burden of data collection, although the protocols used by clinical teams remained consistent with guidelines. The authors discuss the trade-offs inherent in this choice and other decisions made to minimize disruption to usual practice.⁵

Second, this is the largest trial of an inhaled sedation strategy during intensive care and it was designed to assess the impact of this strategy on a patient-centered, clinically meaningful outcome—mortality.⁴ The trial enrolled patients at both academic and community centers with varying levels of staffing and resources, supporting the generalizability of the findings.⁴

Third, this trial reveals a wide variation in intravenous sedation administration, even when inhaled anesthesia is the trial-assigned strategy. Use of concomitant intravenous sedation in the intervention group on any day in the first 14 days after randomization varied across sites from 5.2% to 100%. This variation may result from differences in case mix, clinical familiarity, or practice variation across centers. Regardless of the cause, this variation, which has been described in other trials,⁶ needs to be considered carefully when designing future trials addressing similar questions.

Although this trial did not find evidence of a reduction in mortality from an inhaled strategy with optional intravenous agents vs intravenous sedation alone, it represents an important contribution with its pragmatic design, large size, and broad case mix. Rather than closing the door on this topic, these findings help to refine understanding of the role of inhaled sedation in the ICU and provide an important foundation for future research.