Editorial 

May 19, 2025

The Search for the Single Best Sedative

Edward T. Qian, Todd W. Rice

JAMA. Published online May 19, 2025. doi:10.1001/jama.2025.7359

Critically ill patients receiving invasive mechanical ventilation often require continuous infusion of sedation to address pain and anxiety. The search for the single best sedative agent for critically ill patients has been the topic of investigations for decades. Propofol, a lipid-based γ-aminobutyric acid receptor potentiator, is a commonly used agent in the intensive care unit for all levels of sedation.

Dexmedetomidine is another commonly used sedative in critically ill patients. The agonism of the α₂-adrenergic receptors results in less deep sedation and the α₂-adrenergic receptor agonists are often paired with other adjunctive agents when deeper levels of sedation are desired. Both propofol and dexmedetomidine can have significant adverse effects. Both frequently precipitate hypotension, although propofol is likely at a higher frequency. Dexmedetomidine is well-known to cause bradycardia, which is often dose limiting. Patients who require a longer duration or high doses of propofol are at risk of developing propofol-related infusion syndrome, which is characterized by refractory metabolic acidosis, cardiac dysfunction, hyperlipidemia, and rhabdomyolysis. Propofol-related infusion syndrome carries a mortality rate estimated as high as 37%.¹

Continuous infusions, such as dexmedetomidine and propofol, are commonly used to sedate critically ill patients, and studies^2,3 have demonstrated that best practice is to minimize and regularly interrupt sedation to find the minimally effective dose either through daily cessation until awake or titration to a specific threshold of sedation that is not too deep. Multiple clinical trials have compared dexmedetomidine with propofol in the care of critically ill patients receiving invasive mechanical ventilation. Although patients randomized to receive dexmedetomidine were more cooperative, arousable, and able to communicate pain, the studies^4-6 found similar important clinical outcomes such as mortality, ventilator-free days, delirium-free days, and coma-free days compared with sedation with propofol. Clonidine is another α₂-adrenergic receptor agonist with a similar pharmacotherapeutic profile to dexmedetomidine but with enteral and oral administration; intermittent sedation through the use of oral medications has been suggested as a means to minimize sedation while continuing patient comfort.

In this issue of JAMA, Walsh and colleagues⁷ report the results of the A2B trial, a multicenter, open-label, 3-group randomized clinical trial comparing the use of dexmedetomidine- or clonidine-based sedation vs propofol-based sedation in critically ill adults receiving invasive mechanical ventilation and expected to continue invasive mechanical ventilation for at least 48 hours. Clinicians were allowed to use adjunctive propofol for patients randomized to receive α₂-adrenergic receptor agonists if they needed additional sedation despite reaching the maximum dose for dexmedetomidine or clonidine or if the patient was experiencing dose-limiting adverse effects.

Although guidance was provided for choice of sedation target with dexmedetomidine and clonidine, the final decisions on the dose and duration of sedation, mechanical ventilation weaning, and timing of extubation were ultimately left to the discretion of the clinical teams. These study features, including trying to minimize sedation as much as possible, likely increases the generalizability of the results. The investigators hypothesized that sedation with an α₂-adrenergic receptor agonist would decrease the time from randomization to successful extubation (defined as extubation followed by 48 hours without mechanical ventilation). The secondary outcomes included 180-day mortality, delirium, and bradycardia.

The investigators enrolled more than 1400 patients from 41 intensive care units in the UK between December 2018 and October 2023 with a 6-month pause during the COVID-19 pandemic. Patients were well balanced at baseline among the 3 groups with a median non–neurological Sequential Organ Failure Assessment score of 8 and almost two-thirds of patients presenting with sepsis. Even though receipt of propofol and an opiate for analgesia and sedation was an inclusion criterion, only 1% of patients received dexmedetomidine and 2% of patients received clonidine prior to enrollment.

The study found no difference in time to successful extubation for dexmedetomidine vs propofol (hazard ratio, 1.09 [95% CI, 0.96-1.25], P = .20) or for clonidine vs propofol (hazard ratio, 1.05 [95% CI, 0.95-1.17], P = .34). Similarly, there was no difference in 180-day rates of mortality or delirium. Patients randomized to an α₂-adrenergic receptor agonist experienced a higher prevalence of severe bradycardia (33% in the dexmedetomidine group and 33% in the clonidine group) compared with the propofol group (20%); however, this rate is considerably higher than in other studies of dexmedetomidine.^5,8 The investigators⁷ also found increased rates of agitation in patients randomized to an α₂-adrenergic receptor agonist (dexmedetomidine or clonidine), although it is not clear if this is an effect of patients being more awake and communicative with these agents compared with propofol because the use of rescue medication to treat agitation was similar among the groups.

This large trial of sedation choice in the intensive care unit, and the first to evaluate the use of an oral agent vs continuous sedation, has many strengths. First, the trial design and statistical methods were rigorous and informed by prior literature in the field. Second, the clinical protocols were detailed and the trial methods clear. Third, the groups were largely similar at baseline (although the dexmedetomidine group did contain a slightly higher proportion of younger patients). Fourth, the pragmatic design, inclusion of multiple sedative practices, and enrollment at multiple centers increase the generalizability of the results. Fifth, the results were robust through a number of prespecified sensitivity analyses.

However, there are a few concerns that need to be considered in order to fully interpret the results. The primary outcome of time to successful extubation is patient-centered and logical in a randomized clinical trial comparing sedation strategies in patients receiving invasive mechanical ventilation. However, the trial was unblinded, which might introduce postrandomization differences in care among the groups because many clinicians would not likely feel comfortable extubating with a continuous infusion of propofol (which causes respiratory depression) but may extubate while the patient is receiving continuous or intermittent α₂-adrenergic receptor agonists. Furthermore, the postextubation respiratory support protocols were not clearly defined or driven by the trial protocol. The supplementary material states that spontaneous breaths while receiving noninvasive ventilation were not considered “successful extubation.” Although multiple clinical guidelines^9-11 recommend extubation to noninvasive ventilation routinely for certain patient populations, the guidelines may be inconsistently applied in practice and it is not clear if the practice of extubating to noninvasive ventilation was applied similarly across all 3 trial groups.

Further complicating interpretation of the results is the use of sedation strategies and nonexclusive use of sedation agents. Patients randomized to the dexmedetomidine and clonidine groups received propofol infusions on more than 75% of mechanical ventilation days, albeit at lower doses than the propofol group. This adjunctive use of propofol is pragmatic in that this is how a sedation strategy prioritizing α₂-adrenergic receptor agonists might need to be implemented in clinical care, especially when patients require deeper levels of sedation. However, one of the ways that α₂-adrenergic receptor agonists may decrease the time to successful extubation is by allowing lighter sedation. Unfortunately, multiple secondary measures (including the Richmond Agitation-Sedation Scale and time to first day without any unnecessary deep sedation) were similar among the 3 groups, suggesting that patients in this study may have experienced deeper than intended levels of sedation.

So where does that leave clinicians caring for patients requiring sedation for invasive mechanical ventilation? Unlike the Sevoflurane for Sedation in ARDS (SESAR) trial¹² that found higher mortality and worse clinical outcomes in patients receiving invasive mechanical ventilation randomized to inhaled sevoflurane compared with propofol, the study by Walsh and colleagues⁷ suggests that dexmedetomidine, enteral clonidine, and propofol sedation strategies result in similar clinical outcomes without obvious differences across easily clinically identified subgroups. In the spectrum of sedatives, these results confirm previous results suggesting that no one sedative is best for all critically ill patients.

In-depth phenotyping of patients might delineate subpopulations of patients who benefit from γ-aminobutyric acid potentiation vs α₂-adrenergic receptor agonism.¹³ Until those analyses are complete, the current evidence suggests that clinicians have choices and should use their best judgment with respect to their choice of sedating agent. For example, clinicians may favor propofol if they desire deeper sedation or α₂-adrenergic receptor agonists if they prefer lighter sedation. Bradycardia might result in avoidance of α₂-adrenergic receptor agonist–based sedation. If enteral sedation is desired, clonidine appears to be a reasonable option, especially if trying to avoid using benzodiazepines and antipsychotics. Even though the study by Walsh and colleagues⁷ provides reassurance that currently available, commonly used sedation strategies do not produce big differences in short-term clinical outcomes, further research is needed on the effects of these strategies on longer-term outcomes.

The trial by Walsh and colleagues⁷ suggests that currently there may not be a single best sedative agent. How α₂-adrenergic receptor agonists or propofol might compare with alternative sedation strategies (such as ketamine or propranolol¹⁴) remains largely unknown. Ultimately, the search for the single best sedative agent may be a hopeless endeavor because critically ill patients might benefit from a dynamic sedation strategy that prioritizes different agents at different phases of a patient’s critical illness.