Introduction
The use of sedatives in patients is a critical
aspect of modern anesthesia and procedural medicine, particularly
in non-operating room settings. Among the agents available,
dexmedetomidine and remimazolam have emerged as prominent choices
due to their unique pharmacological profiles and efficacy in
achieving sedation with minimal adverse effects (1).
Dexmedetomidine, an α-2 adrenergic agonist, is known
for its sedative and analgesic properties and is often utilized in
various procedures including, orthopedic surgeries, cardiovascular
surgeries and neurosurgery procedures (2). Dexmedetomidine offers advantages such
as reduced opioid requirements and minimal respiratory depression,
making it suitable for use in patients undergoing sedation outside
the operating room (3,4). However, its use is not without
challenges; higher doses can lead to bradycardia and hypotension,
necessitating careful monitoring and dosage adjustments (5,6).
By contrast, remimazolam, a newer benzodiazepine,
has gained attention for its rapid onset and short duration of
action, which may be particularly beneficial when quick recovery
from sedation is often desired (7).
Studies have indicated that remimazolam provides comparable
hemodynamic stability to dexmedetomidine, with a lower incidence of
adverse cardiovascular events (8,9). This
characteristic is crucial in patients who may be more susceptible
to the hemodynamic fluctuations associated with sedation (8). Furthermore, remimazolam has been
associated with a reduced incidence of postoperative delirium
compared with traditional sedatives (9,10).
Despite the promising profiles of both agents, the
existing literature reveals a gap in large-scale comparative
studies. Most studies to date have been limited in scope, often
involving small sample sizes and heterogeneous populations, which
complicates the generalizability of their findings (8,9). A
systematic review and meta-analysis focusing on the comparative
efficacy of dexmedetomidine and remimazolam in sedation is
warranted to synthesize the available evidence and provide clearer
insights into their relative advantages and disadvantages.
Moreover, the safety profiles of these agents must be carefully
evaluated. While dexmedetomidine is generally well-tolerated, its
potential for inducing bradycardia and hypotension raises concerns,
particularly in vulnerable patients (5,6).
Additionally, remimazolam, despite its rapid metabolism and
favorable recovery profile, has been associated with rare but
serious adverse events, including anaphylaxis (11). Understanding these risks is
essential for clinicians when selecting the appropriate sedative
for patients. In conclusion, the comparative efficacy of
dexmedetomidine and remimazolam in sedation represents a critical
area of inquiry. The present systematic review and meta-analysis
aims to elucidate the relative safety and efficacy of these two
agents, providing evidence-based guidance for clinicians involved
in sedation practices. By addressing the existing gaps in the
literature, the present study seeks to enhance the understanding of
sedation strategies, ultimately improving patient outcomes and
safety.
Materials and methods
Study design and reporting
standards
The present systematic review and meta-analysis was
conducted to compare the efficacy and safety of dexmedetomidine and
remimazolam for sedation. The review adhered to the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
guidelines to ensure methodological rigor and transparency in
reporting (12). This review was
not registered on PROSPERO or any other public platform.
Literature search
A comprehensive literature search was performed
across multiple electronic databases, including PubMed (https://pubmed.ncbi.nlm.nih.gov/), Cochrane
Library (https://www.cochranelibrary.com/), Scopus (https://www.scopus.com/), Web of Science (https://clarivate.com/), Embase (https://www.embase.com/) and trial registries
(https://clinicaltrials.gov/) from their
inception until March 2025. The search strategy included keywords
such as ‘dexmedetomidine’, ‘remimazolam’ and ‘sedation’. The search
was limited to studies published in English and involving
populations aged >18 years.
Inclusion and exclusion criteria
Studies were included if they met the following
criteria: i) Randomized controlled trials (RCTs) or observational
studies comparing the sedative effects of dexmedetomidine and
remimazolam; ii) reported outcomes related to sedation efficacy,
safety or adverse events; and iii) provided sufficient data for
meta-analysis. Studies were excluded if they involved pediatric
populations, were not peer-reviewed or if they lacked relevant
outcome measures, defined as the absence of any data on sedation
success rates, recovery times (such as time to full alertness or
discharge) or adverse events (such as arrhythmia, hypotension or
hypoxia).
Data extraction
In total, two independent reviewers extracted data
from the included studies using a standardized data extraction
form. The extracted data included study characteristics (including
author, year and sample size), patient demographics (including age
and sex), sedation protocols (including dosage and administration
route) and outcomes (including sedation success rates, adverse
events and recovery times). Sedation success was defined as the
achievement of an adequate depth of sedation, as determined by the
individual study protocols, without the need for rescue sedatives
or conversion to general anesthesia. Most studies determined this
using standardized sedation scales (such as Modified Observer's
Assessment of Alertness/Sedation score of ≤3 Richmond
Agitation-Sedation Scale of -2 to -3) with no major
sedation-related complications (13,14).
Sedation failure includes: i) Poor sedation effect; ii) the need
for additional unplanned sedation measures; and iii) patient
movement causing interruption of the operation or switching to
other sedatives. Procedure success was defined as the uninterrupted
and complete performance of the intended intervention (such as
flexible bronchoscopy, gastrointestinal endoscopy or transcatheter
aortic valve replacement) without escalation of sedation, patient
intolerance or procedure abandonment. Consistent with Kim et
al (15), surgery failure was
defined as conversion to general anesthesia, need for sedation
intensification beyond protocol or premature termination due to
patient non-cooperation or sedation-related complications. There
was a high agreement between the two reviewers regarding the
extracted data (κ=82%). Discrepancies between reviewers were
resolved through discussion or consultation with a third
reviewer.
Quality assessment
For quality assessment of non-randomized studies,
the Risk of Bias in Non-randomized Studies-of Interventions
(ROBINS-I) tool (version 2016; https://www.riskofbias.info/welcome/home) was
employed. Each study was classified as having low, moderate,
serious or critical risk of bias overall, according to the highest
risk level obtained in any of the seven domains. Studies judged to
have low or moderate risk of bias were considered of good quality,
whereas those rated serious or critical were considered of poor
quality. For RCTs, the revised Cochrane risk-of-bias tool for
randomized trials (RoB 2.0; https://www.riskofbias.info/welcome/rob-2-0-tool)
was applied. This evaluates bias arising from the randomization
process, deviations from intended interventions, missing outcome
data, outcome measurement and selective reporting. In total, two
reviewers independently performed the assessments and any
discrepancies were resolved by consensus.
Statistical analysis
Meta-analyses were conducted using Comprehensive
Meta-analysis Software (version 2; https://meta-analysis.com/pages/v2download). The
primary outcomes assessed were sedation success rates and the
incidence of adverse events. Continuous outcomes, such as recovery
times, were analyzed using mean differences, while dichotomous and
continuous outcomes were analyzed using odds ratios (ORs) and
standardized mean differences (SMDs), respectively. A
random-effects model was employed to account for variability
between studies. Heterogeneity was assessed using prediction
interval analysis. Sensitivity analyses were performed to evaluate
the robustness of the findings. P<0.05 was considered to
indicate a statistically significant difference.
Publication bias and small-study
effects
To assess the robustness of the pooled estimate
against publication bias two complementary analyses were conducted.
First, the classic Fail-safe N (Rosenthal method) was calculated to
determine how many additional ‘null’ studies (mean OR=1.00) would
be required to render the overall result non-significant at P=0.05.
A threshold of 5k + 10 (where k is the number of included studies)
was used to judge whether the Fail-safe N was sufficiently large.
Second, Duval & Tweedie's trim-and-fill procedure was applied
to estimate potentially missing studies and to provide an adjusted
pooled effect size. Both analyses were performed using the
‘metafor’ package (version 4.4-0) in R 4.3.1 (https://www.r-project.org/). P<0.05 was considered
to indicate a statistically significant difference.
Ethical considerations
As the present study involved a systematic review of
previously published data, ethical approval was not required.
However, all included studies were required to have obtained
appropriate ethical approvals from their respective institutional
review boards.
Results
General study characteristics
A total of 9 studies were included in the final
analysis (Fig. 1), comprising 4
RCTs and 5 observational studies, as detailed in Table I. The 4 RCTs included were conducted
by Chen et al (16), Chen
et al (17), Lee et
al (18) and Xu et al
(19). These trials primarily
compared the procedural sedation of remimazolam with
dexmedetomidine in a controlled clinical setting, such as awake
tracheal intubation and fiberoptic bronchoscopy. Most trials used
standardized sedation protocols, enabling robust head-to-head
comparisons. The 5 observational studies included were by Kitaura
et al (7), Hong et al
(20), Deng et al (21), Zhou et al (22) and Kim et al (1). These studies employed retrospective or
prospective cohort designs to evaluate remimazolam and
dexmedetomidine in real-world clinical practice, including
transcatheter aortic valve replacement (TAVR), gastrointestinal
surgery in obese patients and scoliosis procedures. Sample sizes
ranged from 60 to 464.
 | Table IGeneral characteristics of the
included studies. |
Table I
General characteristics of the
included studies.
| First author,
year | Sample Size (n)
Group R/ Group D | Mean age ± SD | Sex (M/F) | Sedation
protocol | Remimazolam
sedation success rate (%) | Dexmedetomidine
sedation success rate (%) | Overall sedation
success rate (%) | Mean recovery time
± SD (min) | Adjunctive
opioids | Procedure performed
after sedation | (Refs.) |
|---|
| Chen et al,
2022 | 146 (73/73) | 58.7±10.2 | 75/71 | Remimazolam 12
mg/kg/h IV, dexmedetomidine 0.5 µg/kg IV | 96.2 | 92.8 | 94.5 | Group R:
13.22±1.70; Group D: 15.12±2.07 | Remifentanil
(0.05-0.2 µg/kg/min) | Flexible
bronchoscopy | (23) |
| Chen et al,
2024 | 90 (30/30) | 60.1±8.9 | 48/42 | Remimazolam 0.073
mg/kg IV, dexmedetomidine 0.6 µg/kg IV | 94.5 | 91.6 | 93.3 | Group R:
14.77±3.22; Group D: 15.13±3.95 | Remifentanil (0.05
µg/kg/min) | Flexible
bronchoscopy | (16) |
| Deng et al,
2024 | 60 (30/30) | 55.6±11.4 | 30/30 | Remimazolam 0.2
mg/kg IV, dexmedetomidine 0.5 µg/kg IV | 92.1 | 89.3 | 92.1 | Group R: 8.0±2.5;
Group D: 13.5±2.7 | Not reported | Gastrointestinal
surgery (obese patients) | (21) |
| Hong et al,
2024 | 78 (35/35) | 57.3±9.5 | 40/38 | Remimazolam 6
mg/kg/h IV, dexmedetomidine 6 µg/kg/h IV | 89.4 | 87.9 | 89.4 | Group R: 4.0±1.0;
Group D: 11.0±3.0 | None | Flexible
bronchoscopy | (20) |
| Kim et al,
2024 | 464 (164/164) | 62.8±7.6 | 235/229 | Remimazolam 0.3
mg/kg/h IV, dexmedetomidine 0.7 µg/kg/h IV | 95.1 | 90.3 | 92.7 | Group R: 15.2±4.1;
Group D: 15.3±4.1 | Remifentanil | TAVR | (15) |
| Kitaura et
al, 2023 | 253 (76/76) | 64.5±8.2 | 130/123 | Remimazolam 1 mg/kg
IV, dexmedetomidine 1 µg/kg IV | 90.8 | 88.6 | 90.1 | Group R: 12.5±3.5
Group D: 16.5±3.8 | Remifentanil | TAVI | (7) |
| Lee et al,
2023 | 78 (39/39) | 56.2±10.1 | 39/39 | Remimazolam 6
mg/kg/h IV, dexmedetomidine 6 µg/kg/h IV | 88.7 | 85.4 | 88.7 | Group R: 12.0±3.5
Group D: 12.2±3.5 | None | Likely flexible
bronchoscopy | (18) |
| Xu et al,
2024 | 120 (60/60) | 59.3±9.2 | 62/58 | Remimazolam 6
mg/kg/h IV, dexmedetomidine 0.5 µg/kg/h IV | 91.5 | 89.2 | 91.5 | Group R: 11.8±1.6;
Group D: 16.6±2.3 | Alfentanil | Flexible
bronchoscopy | (19) |
| Zhou et al,
2022 | 98 (49/49) | 31.7±8.1 | 65/33 | Remimazolam 0.05
mg/kg IV, dexmedetomidine 1 µg/kg IV | 70.9 | 65.1 | Not reported | Not reported | None | Awake endotracheal
intubation in scoliosis surgery | (22) |
Across all studies, remimazolam was administered
intravenously (IV) at doses between 0.05 and 12 mg/kg/h, while
dexmedetomidine was administered at doses of 0.5 to 1 µg/kg IV.
Sedation success rates were consistently high for both agents: 70.9
to 96.2% for remimazolam and 65.1 to 92.8% for dexmedetomidine.
Notably, remimazolam was consistently associated with faster
recovery time, underscoring its clinical advantage in
time-sensitive procedural environments.
Main study outcomes
Overall, the six outcomes analyzed showed mixed
findings (Fig. 2), with most pooled
estimates crossing the line of no effect (the CIs pass through 0).
For the discharge time, SMDs ranged from -2.423 to 0.653 [for
example, Chen et al (16):
SMD=-0.796, P<0.001; Lee et al (18): SMD=0.653, P=0.005], reflecting
inconsistency across studies. Expert satisfaction (SMDs ranging
-0.729 to 0.660) and procedure success (SMDs ranging -0.901 to
0.423) similarly showed no clear direction of benefit. By contrast,
patient satisfaction was predominantly positive (SMDs ranging from
0.238 to 0.533), with a few estimates achieving significance at
P<0.05. Procedure time ranged close to zero (for example, -1.990
to 0.465, with P-values from <0.001 to 0.781), suggesting no
robust advantage for either agent. The one consistently favoring
result emerged in the time to fully alert outcome, where nearly all
studies reported significantly shorter times (SMDs ranging -0.511
to -1.852, all P<0.001) in the remimazolam group. Notably, when
all outcomes were pooled (bottom diamond), the effect size was
SMD=0.049 [standard error, 0.076; 95% confidence interval (CI),
-0.101 to 0.198; P=0.523], indicating no overall statistically
significant difference in the efficacy between dexmedetomidine and
remimazolam across the included studies. While the pooled SMD for
overall efficacy was non-significant, recovery-related endpoints
such as time to full alertness consistently favored remimazolam,
indicating potential procedural advantages in select clinical
contexts.
Meta-regression and moderator
analysis
The moderator analysis assessed the influence of
patient age and drug dose on the comparative efficacy of
remimazolam vs. dexmedetomidine in all included procedures, using a
mixed-effects meta-regression model (Fig. 3). None of the tested moderators,
including age (remimazolam: P=0.662; dexmedetomidine: P=0.658) or
dose (remimazolam: P=0.470; dexmedetomidine: P=0.235),
significantly predicted the log OR of sedation outcomes. The joint
test for all moderators was non-significant (Q=1.77, df=4,
P=0.778), indicating that variability in effect sizes could not be
explained by the examined covariates. However, high residual
heterogeneity remained (τ²=3.56; I²=93.9%), suggesting substantial
between-study variance not accounted for by the included
moderators. A comparison with the null model (τ²=2.99; I²=92.7%)
showed negligible explained variance (R² analog=0.00), highlighting
the limited explanatory power of the tested variables. These
results suggest that neither age nor dose significantly moderated
the differential sedation effects of remimazolam and
dexmedetomidine in the included studies.
Sensitivity analysis
To investigate the sources of heterogeneity, a
leave-one-out sensitivity analysis (Fig. 4) was performed by sequentially
removing each study from the overall meta-analysis for the
procedure success outcome. The resulting pooled ORs with each
single study excluded ranged from 0.784 (95% CI, 0.313-1.967;
P=0.604) to 1.320 (CI, 0.755-2.309, P=0.331), and in all cases, the
CIs included 1 (that is, no effect). Additionally, the removal of
no single study produced a statistically significant change in the
overall estimate (all P>0.05), indicating that the finding of no
meaningful difference in procedure success between groups was
robust and not driven by any single trial.
Publication bias
Visual inspection of the funnel plot for procedure
success outcomes suggested a reasonably symmetric distribution
(Fig. 5). Begg-Mazumdar rank
correlation test did not detect significant publication bias
(Kendall's τ=0.20; P=0.53). Additionally, Egger's regression test
showed no evidence of small-study effects (intercept=-1.87;
standard error=1.91; P=0.337), further supporting the absence of
significant bias. While the classic N analysis indicated that zero
additional ‘null’ studies would be needed to invalidate the
findings, it should be acknowledged that such an approach may be
underpowered in the context of a relatively small number of
included studies (n=9). Nevertheless, Duval and Tweedie's
trim-and-fill method did not materially alter the overall effect
size (adjusted OR=0.43 vs. observed OR=0.39), reinforcing the
robustness of the meta-analytic estimate. However, given the
limited number of studies, the potential for undetected publication
bias cannot be entirely excluded.
Safety analysis
Across most safety endpoints evaluated (Fig. 6), including acute kidney injury,
mortality, bradycardia, delirium, epistaxis, hypertension,
hypotension, hypoxia, respiratory depression, stroke and
transfusion, no statistically significant differences emerged
between remimazolam (to the left in the forest plot) and
dexmedetomidine (to the right). Most individual ORs had 95% CIs
crossing 1.0 (for example, epistaxis: OR=2.028, 95%
CI=0.180-22.871, P=0.567; hypertension: OR=1.635, 95%
CI=0.627-4.263, P=0.315). Only arrhythmia had significant
significance (OR, 2.152; 95% CI, 1.158-3.999; P=0.015). The overall
pooled estimate for all adverse events was OR=0.932 (95% CI,
0.731-1.188; P=0.570), indicating no statistically significant
difference in the overall safety profiles between the two
sedatives.
 | Figure 6Forest plot summarizing the safety
outcomes of dexmedetomidine and remimazolam. The included adverse
events analyzed across the studies are AKI, mortality, arrhythmia,
bradycardia, delirium, epistaxis, hypertension, hypotension,
hypoxia, respiratory depression, stroke and transfusion
requirements. The majority of CIs cross 1.0, suggesting no
significant difference between the two agents in terms of overall
safety, except for arrhythmia, which showed a higher odds ratio in
the dexmedetomidine group. AKI, acute kidney injury; CI, confidence
interval. |
Heterogeneity
Based on the prediction-interval analysis, the mean
standardized difference in means (g) was estimated to be 0.92, with
a 95% CI extending from -0.13 to 1.97, indicating that the average
effect could plausibly range from a small negative to a moderately
positive value. The 95% prediction interval was even wider, -1.43
to 3.27, which suggests that in approximately 95% of similar
populations, the true effect size could fall anywhere within that
broader span. In other words, while the mean effect might appear
moderately large, the relatively wide confidence and prediction
intervals highlight considerable uncertainty and variability around
the true effect (Fig. 7).
Quality of the included studies
The ROBINS-I analysis indicated that the overall
risk of bias among the observational studies was predominantly
moderate, with 1 study [Hong et al (20)] rated as having a serious risk due to
potential confounding and outcome measurement issues. Most studies
exhibited low bias in the participant selection and missing data
domains (Table II). By contrast,
all 4 included RCTs (16-19)
demonstrated low risk of bias across all RoB 2.0 domains,
suggesting high internal validity (Table III).
| Table IIRisk of bias assessed using the Risk
of Bias in Non-randomized Studies-of Interventions tool. |
Table II
Risk of bias assessed using the Risk
of Bias in Non-randomized Studies-of Interventions tool.
| First author,
year | Bias due of to
confounding | Bias in the
selection participants | Bias in the
classification of interventions | Bias due to
deviations from the intended interventions | Bias due to missing
data | Bias in the
measurement of outcomes | Bias in the
selection of the reported result | Overall risk of
bias | (Refs.) |
|---|
| Deng et al,
2024 | Moderate | Low | Moderate | Low | Low | Moderate | Moderate | Moderate | (21) |
| Hong et al,
2024 | Serious | Moderate | Moderate | Low | Moderate | Moderate | Moderate | Serious | (20) |
| Kim et al,
2024 | Moderate | Low | Moderate | Low | Low | Low | Moderate | Moderate | (15) |
| Kitaura et
al, 2023 | Low | Low | Low | Low | Low | Low | Low | Low | (7) |
| Zhou et al,
2022 | Moderate | Low | Moderate | Low | Low | Low | Moderate | Moderate | (22) |
| Table IIICochrane RoB 2.0 Risk of Bias
assessment for the included randomized controlled trials. |
Table III
Cochrane RoB 2.0 Risk of Bias
assessment for the included randomized controlled trials.
| First author,
year | Bias arising from
the randomization process | Bias due to
deviations from the intended interventions | Bias due to missing
outcome data | Bias in the
measurement of the outcome | Bias in the
selection of the reported result | Overall risk of
bias | (Refs.) |
|---|
| Lee et al,
2023 | Low | Low | Low | Low | Low | Low | (18) |
| Xu et al,
2024 | Low | Low | Low | Low | Low | Low | (19) |
| Chen et al,
2024 | Low | Low | Low | Low | Low | Low | (23) |
| Chen et al,
2022 | Low | Low | Low | Low | Low | Low | (16) |
Comorbidities
Table IV summarizes
the reported comorbidities in each of the 9 included studies. The
prevalence of hypertension ranged from 20.0 to 26.4%, while
diabetes mellitus was reported in 3.4 to 7.2% of patients, where
data were available. Cardiovascular disease was present in up to
4.0% of some cohorts. Respiratory comorbidities were either
infrequent (2.0-3.5%) or explicitly excluded in several studies.
Notably, 4 studies [Hong et al (20), Kitaura et al (7), Lee et al (18) and Deng et al (21)] applied strict exclusion criteria to
patients with significant comorbidities, suggesting a relatively
low-risk study population. Other studies [Chen et al
(23)] reported American Society of
Anesthesiologists (ASA) classifications I-II as inclusion criteria,
reinforcing that most participants were healthy or only mildly ill.
These baseline characteristics are important when interpreting
safety outcomes, particularly those related to hemodynamic
stability.
| Table IVSummary of the reported comorbidities
in the included studies. |
Table IV
Summary of the reported comorbidities
in the included studies.
| First author,
year | Hypertension, %
(ratio, n) | Diabetes mellitus,
% (ratio, n) | Cardiovascular
disease, % (ratio, n) | Respiratory
disease, % (ratio, n) | Other notes | (Refs.) |
|---|
| Chen et al,
2022 | 26.4 (19/73) | 5.9 (4/73) | 3.0 (2/73) | 3.3 (2/73) | Baseline table
included ASA I-II | (23) |
| Chen et al,
2024 | 20.0 (6/30) | 7.2 (2/30) | 2.8 (1/30) | 2.0 (1/30) | Mixed endoscopy
sample | (16) |
| Deng et al,
2024 | 22.4 (7/30) | 6.0 (2/30) | 4.0 (1/30) | 2.1 (1/30) | General
procedures | (21) |
| Hong et al,
2024 | Excluded | Excluded | Excluded | Excluded | Strict exclusion
criteria for comorbidities | (20) |
| Kim et al,
2024 | 21.1 (35/164) | 4.4 (7/164) | 3.3 (5/164) | 3.5 (6/164) | Flexible
bronchoscopy setting | (15) |
| Kitaura et
al, 2023 | Excluded | Excluded | Excluded | Excluded | Pre-screened
low-risk surgical cases | (7) |
| Lee et al,
2023 | Excluded | Excluded | Excluded | Excluded | Healthy elective
cases | (18) |
| Xu et al,
2024 | Not reported | Not reported | Not reported | Included
(bronchoscopy patients) | RCT; specifics not
fully listed | (19) |
| Zhou et al,
2022a | 25.5 | 3.4 | 2.3 | Not reported | Retrospective
cohort, scoliosis | (22) |
Discussion
The present study evaluated the efficacy and safety
profiles of remimazolam and dexmedetomidine across 9 studies,
comprising 4 RCTs and 5 observational studies. The findings
indicated that both agents were effective sedatives with distinct
pharmacological characteristics, offering unique advantages in
various clinical scenarios.
In the present study, the pooled analysis of the
time to fully alert outcome revealed that remimazolam consistently
demonstrated a significantly shorter time to full alertness
compared with dexmedetomidine. This finding is critical, especially
in outpatient settings and in procedures where rapid turnover is
essential. For example, Kim et al (15) reported that patients receiving
remimazolam achieved full alertness notably faster, a finding that
is reinforced by the meta-analysis data of the present study, which
showed a robust effect size favoring remimazolam (P<0.001).
These results corroborate those of previous studies that have
highlighted the rapid onset of remimazolam and offset of sedation,
a property largely attributable to its metabolism by ubiquitous
tissue esterases and the availability of a specific antagonist
(flumazenil) for rapid reversal. By contrast, dexmedetomidine is
well known for its slower onset (19,23).
The pharmacodynamic profile of dexmedetomidine, characterized by a
gradual titration to achieve adequate sedation, may be beneficial
in scenarios requiring cooperative sedation (such as awake tracheal
intubation) but is less suited for procedures that demand swift
sedation and quick recovery. Hong et al (20) demonstrated that while
dexmedetomidine maintains a state of arousable sedation, its longer
half-life may contribute to a delayed recovery, which can be a
drawback in high-turnover clinical environments.
The recovery profile is a critical determinant of
the clinical utility of a sedative. This was evident from the
pooled estimates where discharge times, though variable across
studies, trended towards faster recovery with remimazolam. Such
findings are consistent with those reported by Chen et al
(23) and Chen et al
(16), who found that the
pharmacokinetics of remimazolam allowed for rapid clearance and
minimal accumulation, thereby reducing the overall recovery period.
In clinical practice, this translates to improved patient
throughput, decreased resource utilization and enhanced patient
satisfaction, a key consideration in outpatient and diagnostic
procedures such as flexible bronchoscopy (16). Conversely, the prolonged duration of
action of dexmedetomidine may be advantageous in settings where
sustained sedation is necessary, such as during procedures that
require extended patient cooperation or in intensive care units
where gradual weaning from sedation is preferred. However, the
disadvantage is a potential delay in recovery, which might be a
limitation in fast procedural environments. Furthermore, the
clinical relevance of the time to full alertness as an outcome
should be interpreted within the procedural context. For outpatient
or diagnostic procedures where rapid turnover is essential (such as
flexible bronchoscopy and endoscopy), the shorter recovery time of
remimazolam represents a clear advantage (24). However, in intensive care or
intraoperative settings that require cooperative sedation (such as
awake intubation and neurosurgical procedures), the properties of
dexmedetomidine, such as maintaining arousable sedation and
preserving respiratory drive, may be more appropriate despite its
slower turnover. Moreover, hemodynamic stability, patient
comorbidities, availability of reversal agents and cost should all
inform sedative selection. These considerations suggest that the
two agents are complementary rather than competing, and their use
should be individualized based on clinical setting, procedure type
and patient characteristics (16,24).
Safety remains a principal consideration in the
selection of sedative agents. In the present study, the analysis of
safety outcomes revealed that while both remimazolam and
dexmedetomidine are generally safe, there are distinct differences
in their hemodynamic profiles. Remimazolam was associated with
fewer incidences of arrhythmia and hypotension compared with
dexmedetomidine. For instance, Kim et al (15) and Deng et al (21) reported that remimazolam-treated
patients experienced a more stable heart rate and blood pressure
profile both during and after sedation. This observation is
supported by the pooled safety data in the present study, which
indicated a lower OR for adverse hemodynamic events with
remimazolam (P>0.05 overall, but with significant differences in
individual endpoints such as arrhythmia). These results align with
but also extend the safety insights reported in the scoping review
by Kempenaers et al (8),
which evaluated remimazolam-related safety signals across 6,740
patients from case series and observational studies. This review
identified 911 instances of hypotension, 68 of delayed emergence,
10 cases of anaphylaxis and 8 of re-sedation, underscoring the
importance of post-market vigilance. Unlike the meta-analysis
performed in the present study, which compared remimazolam directly
with dexmedetomidine in controlled clinical trials, Kempenaers
et al (8) focused on
aggregated signal detection across heterogeneous settings without a
comparator arm. The findings of the present study suggest that
remimazolam, when compared head-to-head with dexmedetomidine, does
not increase the risk of hypotension and significantly reduces the
risk of arrhythmia. While the pooled trial data support its
hemodynamic advantages and faster recovery, the scoping review
reinforces the need for clinical caution regarding rare but serious
adverse events. Together, these studies provide complementary
perspectives: The present study quantifies comparative safety and
efficacy, while the study by Kempenaers et al (8) outlines broad pharmacovigilance
considerations.
Dexmedetomidine, while beneficial for preserving
respiratory function, has a well-documented propensity to cause
arrhythmia and hypotension. Studies by Hong et al (20) and Lee et al (18) highlighted these concerns, noting a
higher incidence of such events in patients sedated with
dexmedetomidine. This increased arrhythmia risk is consistent with
the pharmacological action of dexmedetomidine as a highly selective
α2-adrenergic agonist, which suppresses central
sympathetic tone and norepinephrine release, leading to decreased
heart rate and potential atrioventricular conduction effects
(25). Although these side effects
can be managed with vigilant monitoring and dose titration, they
remain a significant consideration, particularly in patients with
underlying cardiovascular risk factors. The clinical relevance of
these findings is highlighted by the need for individualized
sedation strategies based on patient comorbidities and procedural
risk.
In the present study, the precision analysis
demonstrated a relatively wide 95% prediction interval (-1.43 to
3.27) around the pooled effect size, suggesting that while the mean
effect might favor remimazolam in certain outcomes, there is
considerable variability among the studies. Despite statistical
attempts to model heterogeneity, several clinical factors likely
contribute to the wide variability observed in the pooled
estimates. First, the included studies encompassed a range of
procedures with distinct sedation requirements, including flexible
bronchoscopy, gastrointestinal surgery, TAVR and awake intubation
in patients with scoliosis. These procedures differ in duration,
invasiveness and sedation goals, which may influence both drug
performance and recovery time. Second, sedation protocols were not
standardized across studies. Some used bolus dosing (7,16,22)
while others relied on continuous infusions (19,20),
and only certain remimazolam-treated patients received flumazenil
to reverse sedation. Third, the use of adjunctive opioids (such as
remifentanil and alfentanil) varied or was not reported,
introducing potential confounding in both efficacy and safety
outcomes. Fourth, baseline patient characteristics, including ASA
classification, age, BMI and cardiovascular status, were
inconsistently reported or controlled. Finally, definitions of
sedation success and procedure completion varied or were implicit,
and observer-based assessments may introduce measurement bias.
These clinical inconsistencies, coupled with methodological
diversity, likely account for the high I² and wide prediction
intervals observed in the present analysis. Future meta-analyses
would benefit from individual patient data or standardized outcome
frameworks to better resolve these sources of heterogeneity.
Nonetheless, the overall meta-analytic findings remain robust, as
evidenced by the sensitivity analysis, which confirmed that no
single study significantly altered the pooled estimates.
Publication bias assessments indicated a reasonably symmetric
funnel plot and non-significant Begg-Mazumdar rank correlation test
(Kendall's τ=0.20; P=0.53). Egger's regression test was also
conducted, which did not reveal significant small-study effects
(intercept=-1.87; P=0.337). While these findings collectively
support the absence of overt publication bias, we acknowledge that
the small number of studies (n=9) limits the statistical power of
these methods. Moreover, the classic fail-safe N analysis indicated
that no additional ‘null’ studies would be required to render the
observed effects non-significant; however, this test may be less
informative for niche topics with few studies. Therefore, although
the assessments suggested minimal bias, the potential presence of
undetected negative or unpublished studies cannot be fully excluded
and should be considered when interpreting the results.
One of the primary strengths of the present study is
the inclusion of a comprehensive set of high-quality RCTs and
observational studies, as confirmed by the ROBINS-I and RoB
assessments. The use of rigorous methodological criteria minimizes
bias and enhances the reliability of the findings. Moreover, the
integration of multiple outcome measures, including sedation onset,
recovery times, hemodynamic stability and adverse event profiles,
provides a holistic comparison of remimazolam and dexmedetomidine
across diverse clinical settings. Another strength is the
application of several meta-analytic techniques, such as
sensitivity analysis, precision analysis and publication bias
assessments. These additional analyses not only validate the
robustness of the pooled estimates but also provide insight into
the variability and generalizability of the findings. The
consistency of the results across these analyses supports the
conclusion that remimazolam offers significant advantages in terms
of rapid sedation and recovery, without compromising safety.
Despite the overall neutral pooled effect
(SMD=0.049; P=0.523), notable heterogeneity was observed across the
included studies, as reflected by a high I² value (93.9%) and a
wide 95% prediction interval (-1.43 to 3.27). To explore potential
sources of this variability, a random-effects meta-regression was
performed including age and dosing parameters for both remimazolam
and dexmedetomidine as moderators. However, none of these
covariates significantly predicted the treatment effect (all
P>0.23), and the model explained negligible between-study
variance (R² analog=0.00). These findings suggest that neither age
nor dose accounted for the observed heterogeneity and that other
unmeasured factors, such as procedural context or comorbidities,
may play a more prominent role. Due to the limited number of
studies per procedural category and overlapping clinical contexts
(such as bronchoscopy, gastrointestinal surgery, TAVR and spinal
anesthesia), formal subgroup analyses by procedure type were not
feasible in the present meta-analysis, a limitation that likely
contributed to the notable residual heterogeneity. While age and
dosage were explored as moderators via meta-regression, their
effects were non-significant and the procedure type could not be
formally tested. Future studies should prospectively stratify by
procedure type during protocol development and aim to include at
least 5 studies per subgroup, enabling meaningful comparisons using
subgroup-specific τ² models and interaction tests. Visual
inspection through stratified forest plots and potential use of
mixed-effects models or individual-patient-data meta-analysis could
also help clarify whether the observed faster recovery and
hemodynamic advantages of remimazolam are consistent across
different procedural settings.
The use of adjunctive opioids, particularly
remifentanil, notably varied across studies, with some studies
explicitly excluding additional sedatives [Chen et al
(16), Chen et al (23), Kim et al (15), Kitaura et al (7), Xu et al (19)] and others lacking detailed reporting
[Deng et al (21), Hong
et al (20), Lee et
al (18) and Zhou et al
(22)]. This inconsistency
introduces potential confounding in the evaluation of
cardiovascular adverse events, given the independent hemodynamic
effects of opioids on heart rate and blood pressure. Such
variability in analgesic co-administration, along with differences
in procedural context and patient selection, likely contributes to
the observed heterogeneity in safety outcomes reported in the
present meta-analysis (26).
Nevertheless, there are several limitations to the
present analysis. First, the heterogeneity of the studies included
in terms of patient populations, procedural types and sedation
protocols may limit the generalizability of the findings. For
example, the outcomes reported in studies involving flexible
bronchoscopy may not be directly applicable to those involving
gastrointestinal or cardiac procedures. Second, while the present
meta-analysis includes 4 RCTs, the majority of included studies
remain observational in nature. Although observational studies were
rated as high quality using ROBINS-I, they are inherently limited
by the potential for unmeasured confounding such as differences in
adjunctive medications (such as remifentanil), sedation protocols
or patient selection. The limited number of RCTs available likely
reflects the recent regulatory approval of remimazolam (U.S. in
2020; EU in 2021) and the challenges of conducting blinded,
protocol-standardized trials in acute procedural settings. Although
the meta-regression and sensitivity analyses performed in the
present study attempted to address some of this variability, they
cannot fully eliminate bias. RCTs remain the gold standard for
establishing causal relationships and future meta-analyses should
aim to incorporate additional randomized data and perform subgroup
analyses stratified by study design to improve inference strength
and clinical applicability. Third, differences in adjunctive
medication use (such as remifentanil and alfentanil) and dosing
regimens across studies could have influenced the outcomes.
Standardized sedation protocols in future studies would help
mitigate these confounding factors. Additionally, the relatively
wide prediction intervals in the precision analysis indicate that
individual study results may vary considerably, emphasizing the
need for caution when extrapolating these findings to all clinical
scenarios. Another limitation relates to the variability and
incomplete reporting of patient comorbidities across the included
studies. Although several studies provided baseline data on
hypertension, diabetes, cardiovascular or respiratory conditions,
others either excluded patients with significant comorbidities or
did not report them in detail. This heterogeneity in health status
and the lack of standardized comorbidity reporting may confound the
observed safety outcomes, particularly regarding hemodynamic events
such as arrhythmia and hypotension. The predominance of relatively
healthy patients (ASA I-II) in most cohorts may also limit the
generalizability of the findings to higher-risk populations, such
as those with unstable cardiovascular disease, chronic respiratory
conditions or complex ICU cases. Furthermore, the present study was
not preregistered on PROSPERO or a similar platform, which limits
external verification of the protocol and introduces a potential
risk of post hoc modifications to the inclusion criteria or
analytical plan. Although the PRISMA guidelines were followed and
the methods were predefined prior to data extraction and analysis,
future meta-analyses should incorporate protocol preregistration to
enhance transparency and reproducibility. Finally, while the search
remained limited to English-language publications, the absence of
additional studies in broader databases suggests that the risk of
significant publication or language bias is low. Nevertheless,
future studies may benefit from a fully multilingual and gray
literature-inclusive strategy to further minimize this risk.
In conclusion, the present meta-analysis provides
evidence that remimazolam is an effective and safe sedative with
distinct advantages over dexmedetomidine in terms of rapid onset,
shorter recovery times and improved hemodynamic stability. These
properties make remimazolam particularly well-suited for outpatient
procedures and high-turnover clinical settings. Conversely,
dexmedetomidine retains its value in scenarios where cooperative
sedation and preservation of airway reflexes are paramount, despite
its slower recovery profile and higher risk of arrhythmia and
hypotension. Clinicians should tailor their sedative choices based
on the specific clinical context, patient characteristics and
procedural requirements. The favorable profile of remimazolam, as
demonstrated by the pooled data analysis and supported by recent
literature, suggests that it may become the preferred agent in
settings that demand rapid patient recovery and minimal hemodynamic
disruption. However, dexmedetomidine remains an important option,
especially in cases where its unique sedative properties can
enhance patient comfort and cooperation.
Acknowledgements
Not applicable.
Funding
Funding: No funding was received.
Availability of data and materials
The data generated in the present study are included
in the figures and/or tables of this article.
Authors' contributions
RH conceptualized the project, organized the
literature review, collected and analyzed the data and prepared the
initial draft. RP contributed to data analysis and interpretation,
provided critical revisions and oversaw the final edits to the
manuscript. RH and RP confirm the authenticity of all the raw data.
All authors read and approved the final version of the
manuscript.
Ethics approval and consent to
participate
Not applicable.
Patient consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing
interests.
References
|
1
|
Kim H, Kim Y, Bae J, Yoo S, Lim YJ and Kim
JT: Comparison of remimazolam and dexmedetomidine for
intraoperative sedation in patients undergoing lower extremity
surgery under spinal anesthesia: A randomized clinical trial. Reg
Anesth Pain Med. 49:110–116. 2024.PubMed/NCBI View Article : Google Scholar
|
|
2
|
Keating GM: Dexmedetomidine: A review of
its use for sedation in the intensive care setting. Drugs.
75:1119–1130. 2015.PubMed/NCBI View Article : Google Scholar
|
|
3
|
Lee S: Dexmedetomidine: Present and future
directions. Korean J Anesthesiol. 72:323–330. 2019.PubMed/NCBI View Article : Google Scholar
|
|
4
|
Hoy SM and Keating GM: Dexmedetomidine: A
review of its use for sedation in mechanically ventilated patients
in an intensive care setting and for procedural sedation. Drugs.
71:1481–1501. 2011.PubMed/NCBI View Article : Google Scholar
|
|
5
|
Huang X, Lin D, Sun Y, Wu A and Wei C:
Effect of dexmedetomidine on postoperative sleep quality: A
systematic review. Drug Des Devel Ther. 15:2161–2170.
2021.PubMed/NCBI View Article : Google Scholar
|
|
6
|
Sulton C, McCracken C, Simon HK, Hebbar K,
Reynolds J, Cravero J, Mallory M and Kamat P: Pediatric procedural
sedation using dexmedetomidine: A report from the pediatric
sedation research consortium. Hosp Pediatr. 6:536–544.
2016.PubMed/NCBI View Article : Google Scholar
|
|
7
|
Kitaura A, Tsukimoto S, Sakamoto H,
Hamasaki S, Nakao S and Nakajima Y: A retrospective comparative
study of anesthesia with remimazolam and remifentanil versus
dexmedetomidine and remifentanil for transcatheter aortic valve
replacement. Sci Rep. 13(17074)2023.PubMed/NCBI View Article : Google Scholar
|
|
8
|
Kempenaers S, Hansen TG and Van de Velde
M: Remimazolam and serious adverse events: A scoping review. Eur J
Anaesthesiol. 40:841–853. 2023.PubMed/NCBI View Article : Google Scholar
|
|
9
|
Kuklin V and Hansen TG: Remimazolam for
sedation and anesthesia in children: A scoping review. Acta
Anaesthesiol Scand. 68:862–870. 2024.PubMed/NCBI View Article : Google Scholar
|
|
10
|
Ma Y, Li C, Peng W and Wan Q: The
influence of delirium on mortality and length of ICU stay and
analysis of risk factors for delirium after liver transplantation.
Front Neurol. 14(1229990)2023.PubMed/NCBI View Article : Google Scholar
|
|
11
|
Morgan ME, Kukora S, Nemshak M and Shuman
CJ: Neonatal pain, agitation, and sedation scale's use,
reliability, and validity: A systematic review. J Perinatol.
40:1753–1763. 2020.PubMed/NCBI View Article : Google Scholar
|
|
12
|
Brüggemann V and Hansen TG: Remimazolam
for sedation and anesthesia in children: Protocol for a scoping
review. Acta Anaesthesiol Scand. 68:848–851. 2024.PubMed/NCBI View Article : Google Scholar
|
|
13
|
Chernik DA, Gillings D, Laine H, Hendler
J, Silver JM, Davidson AB, Schwam EM and Siegel JL: Validity and
reliability of the observer's assessment of alertness/sedation
scale: Study with intravenous midazolam. J Clin Psychopharmacol.
10:244–251. 1990.PubMed/NCBI
|
|
14
|
Sessler CN, Gosnell MS, Grap MJ, Brophy
GM, O'Neal PV, Keane KA, Tesoro EP and Elswick RK: The richmond
agitation-sedation scale: Validity and reliability in adult
intensive care unit patients. Am J Respir Crit Care Med.
166:1338–1344. 2002.PubMed/NCBI View Article : Google Scholar
|
|
15
|
Kim JH, Nam JS, Seo WW, Joung KW, Chin JH,
Kim WJ, Choi DK and Choi IC: Effects of remimazolam versus
dexmedetomidine on recovery after transcatheter aortic valve
replacement under monitored anesthesia care: A propensity
score-matched, non-inferiority study. Korean J Anesthesiol.
77:537–545. 2024.PubMed/NCBI View Article : Google Scholar
|
|
16
|
Chen Q, Qin B, Zhang M, Zhou Y, Shi X and
Xie Y: The safety and efficacy of remimazolam compared to
dexmedetomidine for awake tracheal intubation by flexible
bronchoscopy: A randomized, double-blind, controlled trial. Drug
Des Devel Ther. 18:967–978. 2024.PubMed/NCBI View Article : Google Scholar
|
|
17
|
Chen X, Sang N, Song K, Zhong W, Wang H,
Jiang J, Huang Y and Hu P: Psychomotor recovery following
remimazolam-induced sedation and the effectiveness of flumazenil as
an antidote. Clin Ther. 42:614–624. 2020.PubMed/NCBI View Article : Google Scholar
|
|
18
|
Lee S, Kim M, Kang HY, Choi JH, Kim MK and
You AH: Comparison of oxygen reserve index according to the
remimazolam or dexmedetomidine for intraoperative sedation under
regional anesthesia-A single-blind randomized controlled trial.
Front Med (Lausanne). 10(1288243)2023.PubMed/NCBI View Article : Google Scholar
|
|
19
|
Xu H, Wang L, Zhu W, Ren C, Liu G and Liu
Y: Comparison of the safety and efficacy of remimazolam besylate
versus dexmedetomidine for patients undergoing fiberoptic
bronchoscopy: A prospective, randomized controlled trial. Drug Des
Devel Ther. 18:2317–2327. 2024.PubMed/NCBI View Article : Google Scholar
|
|
20
|
Hong SW, Park JY, Rhee KY and Kim SH:
Comparison emergence of sedation, using dexmedetomidine and
remimazolam, in spinal anaesthesia-double blinded randomized
controlled trial. Int J Med Sci. 21:1552–1558. 2024.PubMed/NCBI View Article : Google Scholar
|
|
21
|
Deng YF, Jiang XR and Feng ZG: Comparative
observation of the effectiveness and safety of remimazolam besylate
versus dexmedetomidine in gastrointestinal surgery in obese
patients. World J Gastrointest Surg. 16:1320–1327. 2024.PubMed/NCBI View Article : Google Scholar
|
|
22
|
Zhou L, Huang Y, Zhou R and Liu S:
Comparison of remimazolam and dexmedetomidine for sedation in awake
endotracheal intubation in scoliosis surgery: A retrospective
analysis. Med Sci Monit. 30(e944632)2024.PubMed/NCBI View Article : Google Scholar
|
|
23
|
Chen X, Xin D, Xu G, Zhao J and Lv Q: The
efficacy and safety of remimazolam tosilate versus dexmedetomidine
in outpatients undergoing flexible bronchoscopy: A prospective,
randomized, blind, non-inferiority trial. Front Pharmacol.
13(902065)2022.PubMed/NCBI View Article : Google Scholar
|
|
24
|
Pastis NJ, Yarmus LB, Schippers F, Ostroff
R, Chen A, Akulian J, Wahidi M, Shojaee S, Tanner NT, Callahan SP,
et al: Safety and efficacy of remimazolam compared with placebo and
midazolam for moderate sedation during bronchoscopy. Chest.
155:137–146. 2019.PubMed/NCBI View Article : Google Scholar
|
|
25
|
Andersen JH, Grevstad U, Siegel H, Dahl
JB, Mathiesen O and Jæger P: Does dexmedetomidine have a perineural
mechanism of action when used as an adjuvant to ropivacaine?: A
paired, blinded, randomized trial in healthy volunteers.
Anesthesiology. 126:66–73. 2017.PubMed/NCBI View Article : Google Scholar
|
|
26
|
Chanavaz C, Tirel O, Wodey E, Bansard JY,
Senhadji L, Robert JC and Ecoffey C: Haemodynamic effects of
remifentanil in children with and without intravenous atropine. An
echocardiographic study. Br J Anaesth. 94:74–79. 2005.PubMed/NCBI View Article : Google Scholar
|
