Intensive Care Med

DOI 10.1007/s00134-010-1877-6 REVIEW

Jen A. Tan
Kwok M. Ho
Use of dexmedetomidine as a sedative
and analgesic agent in critically ill adult
patients: a meta-analysis

Received: 31 December 2009 Abstract Purpose: To assess the requiring interventions in studies that
Accepted: 9 March 2010 effects of using dexmedetomidine as used both a loading dose and main-
a sedative and analgesic agent on tenance doses [0.7 lg kg-1 h-1
Ó Copyright jointly held by Springer and length of intensive care unit (ICU) [relative risk (RR) 7.30, 95% CI
ESICM 2010 1.73–30.81, P = 0.007]. Risks of
stay, duration of mechanical ventila-
tion, risk of bradycardia, and hypotension requiring interventions
hypotension in critically ill adult (RR 1.43, 95% CI 0.78–2.6,
patients. Methods: Two researchers P = 0.25), delirium (RR 0.79, 95%
searched MEDLINE, EMBASE, and CI 0.56–1.11, P = 0.18), self-extu-
the Cochrane controlled trial register bation, myocardial infarction,
independently for randomized hyperglycemia, atrial fibrillation, and
controlled trials comparing dex- mortality were not significantly
medetomidine with a placebo or an different between dexmedetomidine
alternative sedative agent, without and traditional sedative and analgesic
any language restrictions. Results: agents. Conclusions: Significant
A total of 2,419 critically ill patients heterogeneity existed between the
from 24 trials were subject to meta- pooled studies. The limited evidence
analysis. Dexmedetomidine was suggested that dexmedetomidine
associated with a significant reduction might reduce length of ICU stay in
in length of ICU stay [weighted mean some critically ill patients, but the
J. A. Tan
K. M. Ho ()) difference -0.48 days, 95% confi- risk of bradycardia was significantly
Intensive Care Unit, Royal Perth Hospital, dence interval (CI) -0.18 to higher when both a loading dose and
Perth, WA 6000, Australia -0.78 days, P = 0.002], but not high maintenance doses
e-mail: kwok.ho@health.wa.gov.au duration of mechanical ventilation, ([0.7 lg kg-1 h-1) were used.
Tel.: ?61-8-92241056 when compared with an alternative
Fax: ?61-8-92243668 sedative agent. There was, however, Keywords Bradycardia

significant heterogeneity in these two Hypotension
Intensive care unit

K. M. Ho
School of Population Health, outcomes between the pooled studies. Outcomes
Sedation
University of Western Australia, Dexmedetomidine was associated
Perth, Australia with increased risk of bradycardia

Introduction offset, and without local and systemic adverse effects. So

far, no sedative agent in use has all these ideal properties.
Sedation is often used to improve comfort, reduce anxiety Because of redistribution and accumulation of active
and stress, and facilitate nursing care of critically ill metabolites, many sedative agents, such as opioids and
patients in the intensive care unit (ICU). An ideal sedative benzodiazepines, may have unpredictable and prolonged
agent in the ICU should be cheap, rapid in onset and duration of action in critically ill patients [1, 2].
Furthermore, benzodiazepines and propofol have also agent, such as propofol or benzodiazepines, in critically
been reported to be associated with increased risk of ill adult ([18 years old) patients were included. Studies
delirium [3]. Two new sedative and analgesic agents, that used dexmedetomidine for anesthesia in the oper-
remifentanil and dexmedetomidine, have been increas- ating theater but without continuing its use for sedation
ingly used in many ICUs in the past decade. Despite the in the ICU for more than 6 h were excluded. Two
short duration of action, our recent meta-analysis could reviewers examined all identified trials to confirm they
not confirm the potential advantages of remifentanil over fulfilled the inclusion criteria, and recorded the trial
other traditional sedative agents in reducing delirium, characteristics and outcomes independently, using a
duration of mechanical ventilation, and length of ICU stay predesigned data abstraction form. This abstraction form
[4]. was used to record information regarding the quality of
Dexmedetomidine is a highly selective a2-adrenergic the trial such as allocation concealment, randomization
receptor agonist, which binds to transmembrane G pro- method, blinding of treatment, and inclusion and
tein-binding adrenoreceptors in the periphery (a2A), and exclusion criteria. The grading of allocation was based
brain and spinal cord (a2B, a2C) tissues [5]. In contrast to on the Cochrane approach, that is, adequate, uncertain or
other sedative agents, dexmedetomidine also has potential clearly inadequate. All publications of a study were
analgesic effect [6], and may induce a sedative state retrieved and data extracted. There was no disagreement
similar to physiologic sleep without respiratory depres- between the two independent reviewers in the data
sion by acting on a2 receptors in the locus caeruleus abstracted.
[7–9].
Although dexmedetomidine has some unique phar-
macodynamic properties, its benefits and risks in Statistical analysis
critically ill patients remain uncertain. In particular,
cases of severe bradycardia and unexpected deaths have Length of ICU stay, duration of mechanical ventilation,
been reported [10]. Furthermore, its clearance is also and risk of bradycardia and hypotension requiring inter-
reduced by a reduction in cardiac output and hepatic ventions (e.g. atropine, temporary pacing, inotropes,
blood flow, potentially increasing its duration of action vasopressor, intravenous fluids, cessation or reduction in
in critically ill patients [11]. We hypothesized that dosage of the study drug) were the main outcomes of this
dexmedetomidine may be safe, and able to shorten the study. Other outcomes included risk of delirium, self-
length of ICU stay and duration of mechanical ventila- extubation, atrial fibrillation, nausea and vomiting, myo-
tion, when compared with traditional sedative agents in cardial infarction, hyperglycemia, length of hospital stay,
critically ill patients. and hospital mortality.
During the analyses, studies were stratified into using
dexmedetomidine as a sedative and analgesic agent after
elective high-risk surgery (e.g. cardiac, vascular, and
Methods other surgery requiring ICU admission postoperatively)
and in nonelective critically ill patients. In assessing the
Two researchers searched the Cochrane controlled trial risk of bradycardia, studies were stratified into those
register (2009, issue 4), and the EMBASE (January 1988 to which used loading dose and infusion rate greater than
17 December 2009) and MEDLINE databases (1966 to 17 0.7 lg kg-1 h-1 and those which did not use loading
December 2009) independently. During the electronic dose or infusion rate greater than 0.7 lg kg-1 h-1. The
database search, the following exploded medical subject difference between the two strata was assessed by a ratio
heading (MeSH) terms were used: ‘‘dexmedetomidine’’ of relative risks and difference between treatment means
with ‘‘analgesia,’’ ‘‘sedation’’ or ‘‘sedative agent’’ and for categorical and continuous outcomes, respectively
‘‘critically ill,’’ ‘‘trauma,’’ ‘‘sepsis,’’ ‘‘burns,’’ ‘‘ventila- [12]. Categorical outcomes are reported as relative risk
tion,’’ ‘‘intensive care,’’ ‘‘critical care,’’ ‘‘cardiac surgery,’’ (RR) with 95% confidence interval (CI), using a random-
‘‘heart surgery,’’ ‘‘valvular surgery’’ or ‘‘post-operative.’’ effects model; continuous outcomes are reported as
The search was limited to clinical trials, letters, editorial weighted mean difference (WMD), also using a random-
reviews, and randomized controlled trials, without any effects model.
language restrictions. The reference lists of related edito- The presence of heterogeneity between trials was
rials, reviews, and original articles identified were searched assessed by the chi-square statistic, and the extent of
for relevant trials. Finally, the websites of the International inconsistency was assessed using the I2 statistic [13].
Network of Agencies of Health Technology Assessment in I2 [40% was regarded as indicating significant hetero-
Health Care were searched to ensure that all suitable trials geneity in this study. Publication bias was assessed by
were included. funnel plot using hypotension as an endpoint. Sensitivity
Only randomized controlled trials comparing dex- analyses on the association between dexmedetomidine
medetomidine with a placebo or an alternative sedative and length of ICU stay were conducted by excluding
studies that did not have both adequate allocation con- Results
cealment and double blinding, when dexmedetomidine
was only used as a supplement instead of primary sedative Study characteristics
agent by comparing dexmedetomidine with a placebo, or
were funded by the manufacturer of dexmedetomidine A total of 24 trials involving 2,419 critically ill patients
other than providing the study drug. The latter analysis from over 11 countries were identified and subjected to
was conducted to assess whether commercial funding meta-analysis (Fig. 1) [8, 14–35]. Three studies had more
could have created a bias in the results. than one publication, and the data of all publications of
Data were analyzed by using Review Manager (ver- each study were extracted and analyzed as one study in
sion 4.2.6 for Windows; Oxford, UK: The Cochrane this meta-analysis. All except two studies were published
Collaboration, 2003) and a P value of less than 0.05 was in English [16, 35]. Fifteen studies studied patients after
regarded as significant in this meta-analysis. high-risk elective surgery such as cardiac and vascular

Fig. 1 Flowchart showing

study inclusion and exclusion in Trials compared dexmedetomidine with a placebo, analgesic, or hypnotic agent in
this meta-analysis critically ill patients from MEDLINE, EMBASE, and Cochrane Controlled Trials
Register Databases before 17th December 2009 (n = 32)

Trials excluded (n = 8)
- Without reporting the defined endpoints (n = 2)
- Reporting data of other studies that were already included in
this meta analysis (n = 3)
- Used dexmedetomidine during surgery but not in the ICU
(n = 1)
- Not a randomized controlled study (n=1)
- Used dexmedetomidine less than 6 hours in ICU (n=1)

Trials included for detailed data extraction (n=24, a total of 2419 patients)

Trials with information on the following outcomes:

- Length of intensive care unit stay (11 studies with a total of 1224 patients)
- Duration of mechanical ventilation (11 studies with a total of 1861 patients)
- Bradycardia requiring intervention (10 studies with a total of 1064 patients)
- Hypotension requiring intervention (12 studies with a total of 1545 patients)
- Delirium (8 studies with a total of 1754 patients)
- Atrial fibrillation (8 studies with a total of 1158 patients)
- Nausea and vomiting (7 studies with a total of 1070 patients)
- Myocardial infarction (4 studies with a total of 466 patients)
- Hyperglycemia (3 studies with a total of 695 patients)
- Self-extubation (3 studies with a total of 212 patients)
- Length of hospital stay (2 studies with a total of 417 patients)
- Hospital mortality (16 studies with a total of 1839 patients)
Table 1 Characteristics of the included studies

Study: country of Inclusion (mean age Interventions and no. Outcomes and level Allocation concealment,
origin, year of of participants) and of patients analyzed of sedation or blinding, intention to treat,
publication exclusion criteria analgesia targeted lost to follow-up
(reference), funding

Wahlander et al., Inclusion: after elective thoracotomy Dexmedetomidine group (n = 14): Mortality, bradycardia, Allocation concealment
USA, 2005 [14], and not ventilated 0.5 lg kg-1 over 20 min loading then hypotension, AF, and MI unclear, double blinded,
funding not Exclusion: heart failure, second- or 0.4 lg kg-1 h-1 VAS: B3 analysis by intention to
described third-degree heart block, liver failure, Control group (n = 14): saline at the treat, all completed the
emergency surgery, \18 years old, same rate study
contraindication to an epidural catheter Concurrent treatment: thoracic epidural
with local anesthetic
Akin et al., Turkey, Inclusion: after major abdominal Dexmedetomidine group (n = 30): Bradycardia, hypotension, and Allocation concealment
2008 [15], funding surgery and not ventilated 0.6 lg kg-1 over 30 min loading then nausea or vomiting inadequate, double
not described Exclusion: heart failure, kidney 0.2 lg kg-1 h-1 SAS: target not defined blinded, analysis by
or liver failure, sepsis, shock, Control group (n = 30): placebo not VAS: B3 intention to treat, all
dementia, coagulation defect described. Concurrent treatment: lumbar completed the study
epidural with local anesthetic ?/- iv
tenoxicam 20 mg
Kaneko, Japan, 2008 Inclusion: carotid endarterectomy Dexmedetomidine group (n = 25): Hypotension, hypertension, and Allocation concealment
[16], funding not and ventilated postoperatively 0.2–0.7 lg kg-1 h-1 nausea or vomiting unclear, no blinding,
described Exclusion: patients with neurological Propofol group (n = 33): RSS: 3–4 analysis not by intention
deficits before surgery 1–3 mg kg-1 h-1 to treat, all completed the
Concurrent treatment: iv diclofenlac study
25 mg
Shehabi et al., Inclusion: elective on-pump CABG, Dexmedetomidine group (n = 152): no Mortality, delirium, bradycardia, Allocation concealment
Australia, 2009 valve, or combined CABG and valve loading, infusion 0.1–0.7 lg kg-1 h-1, AF, nausea or vomiting, duration adequate, double blinded,
[17], only the drug surgery morphine as analgesia if needed of mechanical ventilation, ICU analysis by intention to
was funded by Exclusion: bradycardia, hypotension, Morphine group (n = 147): and hospital stay (median and treat, 2.3% did not
Hospira body weight [150 kg, renal failure, 10–70 lg kg-1 h-1 IQR) complete the study
dementia, seizure or Parkinson disease Concurrent treatment: propofol if MAAS score: 2–4
systolic blood pressure [160 mmHg or
unplanned awakening
Triltsch et al., Inclusion: major surgery required Dexmedetomidine group (n = 15): Mortality, nausea or vomiting, Allocation concealment
Germany, 2002 ventilation [6 h after surgery 6 lg kg-1 over 10 min loading, duration of mechanical unclear, double blinded,
[18], funded in part Exclusion: unstable diabetes mellitus, infusion 0.1–0.7 lg kg-1 h-1 ventilation (median and IQR) analysis by intention to
by Abbott liver failure, heart block, neurotrauma, Control group (n = 15): placebo not Bispectral index: 60–70 treat, 6.7% did not
Laboratories or obesity described complete the study
Concurrent treatment: propofol bolus
0.2 mg kg-1 and 0.5–4 mg kg-1 h-1
infusion, paracetamol and morphine for
pain
Herr et al., USA, Inclusion: elective CABG Dexmedetomidine group (n = 148): AF, MI, delirium (or agitation), Allocation concealment
2003 [19], funded Exclusion: neurological disease, obesity, 1 lg kg-1 over 20 min loading, nausea or vomiting, hypotension, adequate, not blinded,
by Abbott heart failure, drug overdose, significant infusion 0.2–0.7 lg kg-1 h-1, duration of mechanical analysis by intention to
Laboratories requirement of dobutamine propofol for sedation if optimal ventilation (median and treat, all completed the
([8 lg kg-1 min-1) or noradrenaline sedation was not achieved by IQR) RSS: C3 while intubated study
([4 lg min-1) maximum dose of dexmedetomidine
Propofol group (n = 147): propofol
infusion rate was not standardized
Concurrent treatment: morphine and
NSAID for pain in both groups
Table 1 continued
Study: country of Inclusion (mean age Interventions and no. Outcomes and level Allocation concealment,
origin, year of of participants) and of patients analyzed of sedation or blinding, intention to treat,
publication exclusion criteria analgesia targeted lost to follow-up
(reference), funding

Martin et al., USA Inclusion: major surgery and required Dexmedetomidine group (n = 203): Mortality, delirium, AF, nausea or Allocation concealment
and European ventilation [6 h after surgery 1 lg kg-1 over 10 min loading, vomiting, hypotension, duration unclear, double blinded,
countries, 2003 Exclusion: neurological disease, obesity, infusion 0.2–0.7 lg kg-1 h-1 of mechanical ventilation (mean analysis by intention to
[20], funded by unstable diabetes mellitus, drug Control group (n = 198): normal saline and SD) treat, 2.5% did not
Abbott overdoses, excessive bleeding requiring Concurrent treatment: propofol for RSS: C3 while intubated complete the study
Laboratories reoperation sedation and morphine for pain
Memis et al., Turkey, Inclusion: critically ill patients with Dexmedetomidine group (n = 20): Mortality Allocation concealment
2007 [21], funding bacteriologically confirmed sepsis 1 lg kg-1 over 10 min loading, RSS: \2 adequate, unclear
not described Exclusion: unstable hemodynamics, infusion 0.2–2.5 lg kg-1 h-1 blinding, analysis by
heart, liver or renal failure, and allergic
Midazolam group (n = 20): 0.2 mg kg-1 intention to treat, all
to study drugs loading, 0.1–0.5 mg kg-1 h-1 completed the study
Concurrent treatment: alfentanil infusion
for pain
Memis et al., Turkey, Inclusion: critically ill patients who were Dexmedetomidine group (n = 12): Bradycardia, hypotension, and Allocation concealment
2006 [22], funding enterally fed 2.5 lg kg-1 over 10 min loading, duration of ICU stay (mean and adequate, double blinded,
not described Exclusion: allergic to paracetamol or infusion 0.2 lg kg-1 h-1 SD) analysis by intention to
with paracetamol overdose Propofol group (n = 12): 2 mg kg-1 h-1 Sedation scale not used treat, all completed the
study
Corbett et al., USA, Inclusion: elective CABG Dexmedetomidine group (n = 43): Mortality, delirium, AF, self- Allocation concealment
2005 [23], funded Exclusion: unstable hemodynamics, 1 lg kg-1 over 15 min loading, extubation, hypotension, adequate, no blinding,
by Society of renal or liver failure, obesity, alcohol or infusion 0.2–0.7 lg kg-1 h-1 duration of mechanical analysis by intention to
Critical Care drug abuse, neurological disease Propofol group (n = 46): 5–75 ventilation (mean and SD), and treat, all completed the
Medicine lg kg-1 min-1 length of ICU stay (median and study
Concurrent treatment: IQR)
morphine for pain and midazolam for RSS: 3–4
breakthrough anxiety
Elbaradie et al., Inclusion: major cancer surgery and Dexmedetomidine group (n = 30): Bradycardia Allocation concealment
Egypt, 2004 [24], required [6 h of mechanical 2.5 lg kg-1 over 10 min loading, RSS: 2–5 unclear, single blinded,
funding not ventilation infusion 0.2–0.5 lg kg-1 h-1 analysis by intention to
described Exclusion: obesity, liver or renal failure, Propofol group (n = 30): 1 mg kg-1 treat, all completed the
neurological disease, unstable diabetes loading, infusion 0.5–1 mg kg-1 h-1 study
mellitus, and corticosteroid treatment Concurrent treatment: fentanyl for pain
within 3 months
Ruokonen et al., Inclusion: mixed medical/surgical ICU Dexmedetomidine group (n = 41): Mortality, delirium, bradycardia, Allocation concealment
Finland and and needed sedation [24 h 0.8 lg kg-1 over 60 min loading, hypotension, duration of unclear, double blinded,
Switzerland, 2009 Exclusion: neurological disease, unstable infusion 0.25–1.4 lg kg-1 h-1 mechanical ventilation (median analysis not by intention
[25], funded by hemodynamics, heart block, hearing or Propofol or midazolam group (n = 44): and range), and length of ICU to treat, all completed the
Orion Pharma visual problem, use of alpha-2 agonist at propofol: 2.4 mg kg-1 over 60 min stay (median and range) study
time of randomization loading, infusion 0.8–4 mg kg-1 h-1 RASS: stratified into 0 to -3 or -4
Midazolam: 1–2 mg loading, infusion
0.12–0.2 mg kg-1 h-1
Table 1 continued
Study: country of Inclusion (mean age Interventions and no. Outcomes and level Allocation concealment,
origin, year of of participants) and of patients analyzed of sedation or blinding, intention to treat,
publication exclusion criteria analgesia targeted lost to follow-up
(reference), funding

Memis et al., Turkey, Inclusion: septic shock but Dexmedetomidine group (n = 20): Mortality Allocation concealment
2009 [26], funded hemodynamically stable 1 lg kg-1 over 10 min loading, RSS: \2 adequate, unblinded,
by University of Exclusion: arrhythmias, myocardial infusion 0.2–2.5 lg kg-1 h-1 analysis by intention to
Trakya Research ischemia, and renal failure Propofol group (n = 20): 1 mg kg-1 treat, all completed the
Grant over 15 min loading, infusion study
1–4 mg kg-1 h-1
Concurrent treatment: alfentanil for pain
Tasdogan et al., Inclusion: ileus surgery and required Dexmedetomidine group (n = 20): Mortality, hypotension, duration of Allocation concealment
Turkey, 2009 [27], mechanical ventilation after surgery 1 lg kg-1 over 10 min loading, mechanical ventilation (median adequate, unblinded,
funded by and met two or more sepsis criteria infusion 0.2–2.5 lg kg-1 h-1 and range), and length of ICU analysis not by intention
University of Exclusion: unstable hemodynamics, liver Propofol group (n = 20): 1 mg kg-1 stay (median and range) to treat, all completed the
Trakya Research or renal failure, and brain death over 15 min loading, infusion RSS: \2 study
Grant 1–3 mg kg-1 h-1
Concurrent treatment: alfentanil for pain
Talke et al., USA, Inclusion: elective vascular surgery and Dexmedetomidine group (n = 18): three AF and myocardial infarction Allocation concealment
1995 [28], funded with or at risk of coronary artery regimens targeting serum Sedation scale not used unclear, double blinded,
by Orion disease concentrations 0.15 ng ml-1, analysis not by intention
Corporation Exclusion: unstable angina, left bundle 0.3 ng ml-1, and 0.45 ng ml-1 to treat, 4% did not
branch block, treated with clonidine or Control group (n = 6): not described complete the study
tricyclic antidepressants Concurrent treatment: morphine for pain
Talke et al., USA, Inclusion: elective vascular surgery Dexmedetomidine group (n = 22): Mortality and hypotension Allocation concealment
2000 [29], funded Exclusion: pregnancy, heart block, 1.2 lg min-1 for 20 min, Sedation scale not used adequate, double blinded,
by Orion treated with clonidine or tricyclic 0.8 lg min-1 for 40 min, analysis by intention to
Corporation antidepressants 0.35 lg min-1 for 4 h, then treat, all completed the
0.15 lg min-1 for 43 h study
Control group (n = 19): normal saline
Concurrent treatment: morphine for pain
Venn et al., UK, 2001 Inclusion: elective complex abdominal Dexmedetomidine group (n = 10): Mortality and hypotension Allocation concealment
[30], funded by or pelvic surgery requiring 8 h of 2.5 lg kg-1 over 10 min loading, RSS: [2 adequate, unblinded,
Abbott mechanical ventilation after surgery 0.2–2.5 lg kg-1 h-1 analysis by intention to
Laboratories Exclusion: pregnancy, liver or renal Propofol group (n = 10): 1 mg kg-1 treat, all completed the
failure, treated with etomidate or loading, 1–3 mg kg h-1 study
corticosteroids, and spinal or epidural Concurrent treatment: alfentanil for pain
anesthesia
Venn et al., UK, 1999 Inclusion: elective general or cardiac Dexmedetomidine group (n = 47): Mortality, AF, nausea or vomiting, Allocation concealment
[8], funded by surgery requiring [6 h of mechanical 1 lg kg-1 over 10 min loading, bradycardia, hypotension, MI, unclear, double blinded,
Abbott ventilation 0.2–0.7 lg kg-1 h-1 duration of mechanical analysis not by intention
Laboratories Exclusion: obesity, neurological disease, Control group (n = 51): normal saline ventilation (mean and SD) to treat, 6.7% did not
drug overdoses, excessive bleeding, and Concurrent treatment: midazolam for RSS: [2 while intubated complete the study
unstable diabetes mellitus sedation and morphine for pain
Table 1 continued
Study: country of Inclusion (mean age Interventions and no. Outcomes and level Allocation concealment,
origin, year of of participants) and of patients analyzed of sedation or blinding, intention to treat,
publication exclusion criteria analgesia targeted lost to follow-up
(reference), funding

Reade et al., Inclusion: critically ill patients who Dexmedetomidine group (n = 10): Mortality, self-extubation, and Allocation concealment
Australia, 2009 required mechanical ventilation 1 lg kg-1 over 20 min loading, length of ICU stay (median and adequate, unblinded,
[31], only the drug because of agitation 0.2–0.7 lg kg-1 h-1 IQR) analysis by intention to
was funded by Exclusion: required ongoing airway Haloperidol group (n = 10): 2.5 mg RASS: 0 treat, all completed the
Hospira protection, an adverse event to loading, 0.5–2 mg h-1 infusion study
haloperidol or alpha-2 agonists, and Concurrent treatment: propofol or
those with a plan to return to operating midazolam for sedation and morphine
theater for pain
Maldonado et al., Inclusion: elective cardiac valve surgery Dexmedetomidine group (n = 40): Mortality, delirium, duration of Allocation concealment
USA, 2009 [32], Exclusion: neurological disease, drug 0.4 lg kg-1 loading, 0.2–0.7 mechanical ventilation (mean adequate, unblinded,
funding not abuse, treated with psychotropic drugs, lg kg-1 h-1 and SD), length of ICU and analysis by intention to
described heart block, and use of circulatory arrest Propofol group (n = 38): 25–50 hospital stay (mean and SD), and treat, 24% did not
during surgery lg kg-1 min-1 costs complete the study
Midazolam group (n = 40): RSS: 2–3
0.5–2 mg h-1
Concurrent treatment: haloperidol and
lorazepam for agitation
Fentanyl and ketorolac for pain
Riker et al., USA, Inclusion: mixed medical and surgical Dexmedetomidine group (n = 244): Mortality, delirium, bradycardia, Allocation concealment
Australia, critically ill patients requiring optional 1 lg kg-1 loading, hypotension, duration of adequate, double blinded,
Argentina, Brazil, mechanical ventilation 3 or more days 0.8 lg kg-1 h-1 mechanical ventilation (median analysis not by intention
New Zealand, 2009 Exclusion: trauma, burns, liver or renal Midazolam group (n = 122): 0.05 and SD), length of ICU stay to treat, 2.4% did not
[33], funded by failure, pregnancy, neurological disease, mg kg-1 loadings, 0.06 mg kg-1 h-1 (median and SD) complete the study
Hospira heart block, unstable angina, and use of Concurrent treatment: open-label RASS: -2 to ?1
epidural or spinal analgesia midazolam for inadequate sedation,
fentanyl for pain, and haloperidol for
agitation
Pandharipande et al., Inclusion: mixed medical and surgical Dexmedetomidine group (n = 52): Mortality, delirium, AF, self- Allocation concealment
USA, 2007 [34], critically ill who required mechanical 0.15–1.5 lg kg-1 h-1 extubation, length of ICU stay adequate, double blinded,
funded by Hospira ventilation [24 h Lorazepam group (n = 51): 1–10 (median and IQR), and costs intention to treat, 2.8%
Exclusion: neurological disease, liver mg h-1 Concurrent treatment: RASS score: determined by did not complete the
failure, alcohol abuse, active myocardial Propofol for inadequate sedation and medical team study
ischemia, heart block, benzodiazepine fentanyl for pain
dependency, pregnancy, and severe
hearing disabilities
Ozkan et al., Turkey, Inclusion: elective CABG Dexmedetomidine group (n = 20): Duration of mechanical ventilation Allocation concealment
2007 [35] funding Exclusion: ejection fraction \35%, 0.2–0.4 lg kg-1 h-1 (mean and SD) and length of unclear, unblinded,
not described serious heart failure, drug dependency, Midazolam group (n = 20): ICU stay (mean and SD) analysis by intention to
antipsychotic drug usage, valvular 0.05–0.1 lg kg-1 h-1 RSS: sedation target not defined treat, all completed the
dysfunction, renal, hepatic, pulmonary, study
or endocrine disorders, myocardial
infarction within 3 days, obesity,
neurological diseases
 surgery [8, 14–20, 23, 24, 28–30, 32, 35], and nine studies

AF atrial fibrillation, CABG coronary artery bypass graft, ICU intensive care unit, MI myocardial infarction, SD standard deviation, IRQ interquartile range, NSAID
nonsteroidal anti-inflammatory drug, HELPP hemolysis elevated liver enzymes and platelets syndrome, RASS Richmond agitation and sedation scale, RSS Ramsay sedation
blinding, intention to treat,

analysis by intention to
treat, all completed the
studied nonelective critically ill patients [21, 22, 25–27,

inadequate, unblinded,
Allocation concealment,

Allocation concealment
31, 33, 36]. Seven studies compared dexmedetomidine
with placebo, and 15 trials compared dexmedetomidine
lost to follow-up

with an alternative hypnotic agent (propofol, midazolam,

or lorazepam). The remaining two studies compared dex-
study medetomidine with either morphine [17] or haloperidol
[31]. A loading dose was used in 19 studies. The maximum
maintenance doses of dexmedetomidine ranged between
0.2 and 2.5 lg kg-1 h-1, and in seven studies the maxi-
mum doses of dexmedetomidine were [0.7 lg kg-1 h-1
[21, 25–27, 30, 34]. Four studies used a loading dose and
also maintenance doses [0.7 lg kg-1 h-1 [8, 25, 30, 33].
The overall quality of the studies was modest.
(median and range)
Duration of ICU stay

Although 13 studies had adequate allocation concealment

Outcomes and level

and 12 studies were double blinded, only five had both

analgesia targeted

adequate allocation concealment and double blinding [17,

of sedation or

22, 29, 33, 34]. Half of the included studies (n = 12, 80%
RSS: 2–3

of trials that disclosed their funding) were, either totally or

in part, funded by a pharmaceutical company. The details
of the included trials are described in Table 1.
0.05 mg kg-1, infusion 0.1 mg kg-1 h-1
Midazolam group (n = 20): loading dose

2 g h-1 for 24 h. Propofol and fentanyl

loading dose 1 lg kg-1 over 20 min,

as rescue sedative and analgesic agent,

nitroprusside infusion to control blood

received magnesium sulfate infusion

Main outcomes
Dexmedetomidine group (n = 20):

Concurrent treatment: all patients

respectively. Nitroglycerin and

There was significant heterogeneity in length of ICU

infusion 0.7 lg kg-1 h-1

(I2 = 83.1%) and duration of mechanical ventilation

(I2 = 74.6%) between the included studies and, when
Interventions and no.

pressure if necessary
of patients analyzed

pooled, dexmedetomidine was associated with a reduction

in length of ICU stay (WMD -0.48 days, 95% CI -0.18
to -0.78 days, P = 0.002) (Fig. 2), but not duration of
mechanical ventilation (Fig. 3). The association between
dexmedetomidine and length of ICU stay was not sig-
score, MAAS motor activity assessment scale, VAS visual analog scale

nificantly different between the two strata of studies

(z = 1.65, P = 0.10).
termination of pregnancy by caesarean

Dexmedetomidine did not appear to increase risk of

cardiac, neurological, renal, endocrine,

bradycardia (RR 1.82, 95% CI 0.66–5.03, P = 0.25)

or hepatic disease, and with HELLP

when all trials were pooled together. The risk of brady-

Exclusion: chronic hypertension,

cardia was, however, significantly higher in studies that

Inclusion: eclampsia requiring

used both a loading dose and high maintenance doses

([0.7 lg kg-1 h-1) (5.8% versus 0.4%; RR 7.30, 95% CI
1.73–30.81, P = 0.007) (Fig. 4) than studies that did not
Inclusion (mean age
of participants) and

use both (RR 0.95, 95% CI 0.39–2.34, P = 0.92) (RR

exclusion criteria

ratio between the two strata = 7.68, 95% CI 1.44–30.20,

z = 2.426, P = 0.015). Dexmedetomidine was not asso-
delivery

syndrome

ciated with increased risk of hypotension requiring

interventions (RR 1.43, 95% CI 0.78–2.6, P = 0.25)
(Fig. 5).
Turkey, 2009 [35]
(reference), funding
Table 1 continued
Study: country of

Other outcomes
Esmaoglu et al.,
origin, year of

funding not
described
publication

Dexmedetomidine was not associated with a significant

reduction in risk of delirium (RR 0.79, 95% CI 0.56–1.11,
P = 0.18, I2 = 71.6%) (Fig. 6), atrial fibrillation (RR
0.95, 95% CI 0.68–1.33, P = 0.77, I2 = 0%), nausea and
Study Dexmedetomidine Control WMD (random) Weight WMD (random)
or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI Year

Elective postoperative patients

Corbett [23]
Ozkan [35]
Maldonado [32]
Maldonado [32]
Shehabi [17]
Subtotal (95% CI)
Test for heterogeneity: Chi² = 14.72, df = 4 (P = 0.005), I² = 72.8%
Test for overall effect: Z = 1.18 (P = 0.24)

Non-elective critically ill patients

Memis [22]
Pandharipande [34]
Esmaoglu [36]
Reade [31]
Riker [33]
Ruokonen [25]
Tasdogan [27]
Subtotal (95% CI)
Test for heterogeneity: Chi² = 27.55, df = 6 (P = 0.0001), I² = 78.2%
Test for overall effect: Z = 1.81 (P = 0.07)

Total (95% CI)

Test for heterogeneity: Chi² = 65.08, df = 11 (P < 0.00001), I² = 83.1%
Test for overall effect: Z = 3.14 (P = 0.002)

-4 -2 0 2 4
Favors dexmedetomidine Favors control

Fig. 2 Effect of dexmedetomidine on length of intensive care unit stay in days

Study Dexmedetomidine Control WMD (random) Weight WMD (random)

or sub-category N Mean (SD) N Mean (SD) 95% CI % 95% CI Year

Elective postoperative patients

Venn [8]
Triltsch [18]
Herr [19]
Martin [20]
Corbett [23]
Ozkan [35]
Maldonado [32]
Maldonado [32]
Shehabi [17]
Subtotal (95% CI)
Test for heterogeneity: Chi² = 13.40, df = 8 (P = 0.10), I² = 40.3%
Test for overall effect: Z = 1.16 (P = 0.25)

Non-elective critically ill patients

Riker [33]
Ruokonen [25]
Tasdogan [27]
Subtotal (95% CI)
Test for heterogeneity: Chi² = 27.09, df = 2 (P < 0.00001), I² = 92.6%
Test for overall effect: Z = 0.62 (P = 0.54)

Total (95% CI)

Test for heterogeneity: Chi² = 43.26, df = 11 (P < 0.00001), I² = 74.6%
Test for overall effect: Z = 0.81 (P = 0.42)

-100 -50 0 50 100

Favors dexmedetomidine Favors control

Fig. 3 Effect of dexmedetomidine on duration of mechanical ventilation in hours

vomiting (RR 1.03, 95% CI 0.66–1.59, P = 0.90, Publication bias and sensitivity analyses
I2 = 29.6%), myocardial infarction (RR 0.62, 95% CI
0.07–5.63, P = 0.67, I2 = 50.1%), hyperglycemia (RR Using hypotension as an endpoint, the funnel plot did not
1.05, 95% CI 0.64–1.71, P = 0.85, I2 = 78.1%), self- suggest the presence of publication bias (Fig. 8).
extubation (RR 1.36, 95% CI 0.31–5.90, P = 0.68, After excluding studies funded by the manufacturer of
I2 = 0), mortality (RR 0.85, 95% CI 0.64–1.13, dexmedetomidine (WMD -0.37 days, 95% CI -0.08
P = 0.26, I2 = 0%) (Fig. 7), and length of hospital stay. to -0.67 days, P = 0.01, I2 = 85.5%) or studies that
There was significant heterogeneity in risk of delirium compared dexmedetomidine with a placebo, the effect of
(I2 = 71.6%) and length of hospital stay (I2 = 68.9%) dexmedetomidine on length of ICU stay remained
between the pooled studies. The effect on delirium did not unchanged. After excluding studies that did not have both
appear to be different between the two strata of studies adequate allocation concealment and double blinding,
(RR ratio in delirium 1.76, 95% CI 0.73–4.26, P = 0.21). dexmedetomidine was not associated with reduction in
Study Dexmedetomidine Control RR (random) Weight RR (random)
or sub-category n/N n/N 95% CI % 95% CI Year

No loading dose and high maintenance doses

Elbaradie [24]
Wahlander [14]
Memis [22]
Pandharipande [34]
Akin [15]
Shehabi [17]
Subtotal (95% CI)
Total events: 23 (Dexmedetomidine), 24 (Control)
Test for heterogeneity: Chi² = 3.50, df = 3 (P = 0.32), I² = 14.3%
Test for overall effect: Z = 0.10 (P = 0.92)

Loading dose and high maintenance doses were used

Venn [8]
Venn [30]
Riker [33]
Ruokonen [25]
Subtotal (95% CI)
Total events: 21 (Dexmedetomidine), 1 (Control)
Test for heterogeneity: Chi² = 0.10, df = 2 (P = 0.95), I² = 0%
Test for overall effect: Z = 2.70 (P = 0.007)

Total (95% CI)

Total events: 44 (Dexmedetomidine), 25 (Control)
Test for heterogeneity: Chi² = 9.60, df = 6 (P = 0.14), I² = 37.5%
Test for overall effect: Z = 1.16 (P = 0.25)

0.01 0.1 1 10 100

Favors dexmedetomidine Favors control

Fig. 4 Effect of dexmedetomidine on risk of bradycardia requiring interventions

Study Dexmedetomidine Control RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 95% CI Year

No loading and high maintenance doses

Talke [29]
Herr [19]
Martin [20]
Wahlander [14]
Memis [22]
Akin [15]
Kaneko [16]
Tasdogan [27]
Subtotal (95% CI)
Total events: 42 (Dexmedetomidine), 36 (Control)
Test for heterogeneity: Chi² = 13.26, df = 6 (P = 0.04), I² = 54.7%
Test for overall effect: Z = 0.45 (P = 0.65)

Loading and high maintenance doses were used

Venn [8]
Venn [30]
Riker [33]
Ruokonen [25]
Subtotal (95% CI)
Total events: 86 (Dexmedetomidine), 34 (Control)
Test for heterogeneity: Chi² = 6.44, df = 2 (P = 0.04), I² = 69.0%
Test for overall effect: Z = 1.03 (P = 0.30)

Total (95% CI)

Total events: 128 (Dexmedetomidine), 70 (Control)
Test for heterogeneity: Chi² = 18.22, df = 9 (P = 0.03), I² = 50.6%
Test for overall effect: Z = 1.16 (P = 0.25)

0.01 0.1 1 10 100

Favors dexmedetomidine Favors control

Fig. 5 Effect of dexmedetomidine on risk of hypotension requiring interventions

length of ICU stay (WMD -0.79 days, 95% CI -2.06 to patients, but the risk of bradycardia was significantly
0.48 days, P = 0.22, I2 = 69.1%). higher when both a loading dose and high maintenance
doses ([0.7 lg kg-1 h-1) were used.
Our results showed that dexmedetomidine might be
able to reduce length of ICU stay in some critically ill
Discussion patients. Although dexmedetomidine is more expensive
than traditional sedative agents, it may still be potentially
This meta-analysis showed that significant heterogeneity cost-effective if it can reduce length of ICU stay [37]. We
between studies on dexmedetomidine existed. The limited should, however, interpret this finding with caution. First,
evidence suggested that dexmedetomidine might be able there was significant heterogeneity between the pooled
to reduce the length of ICU stay in some critically ill studies in this outcome. This heterogeneity could be due
Study Dexmedetomidine Control RR (random) Weight RR (random)
or sub-category n/N n/N 95% CI % 95% CI Year

Elective postoperative patients

Herr [19]
Martin [20]
Corbett [23]
Maldonado [32]
Shehabi [17]
Subtotal (95% CI)
Total events: 25 (Dexmedetomidine), 63 (Control)
Test for heterogeneity: Chi² = 7.66, df = 4 (P = 0.10), I² = 47.8%
Test for overall effect: Z = 1.49 (P = 0.14)

Non-elective critically ill patients

Pandharipande [34]
Riker [33]
Ruokonen [25]
Subtotal (95% CI)
Total events: 191 (Dexmedetomidine), 146 (Control)
Test for heterogeneity: Chi² = 11.91, df = 2 (P = 0.003), I² = 83.2%
Test for overall effect: Z = 0.32 (P = 0.75)

Total (95% CI)

Total events: 216 (Dexmedetomidine), 209 (Control)
Test for heterogeneity: Chi² = 24.61, df = 7 (P = 0.0009), I² = 71.6%
Test for overall effect: Z = 1.36 (P = 0.18)

0.01 0.1 1 10 100

Favors dexmedetomidine Favors control

Fig. 6 Effect of dexmedetomidine on risk of delirium

Study Dexmedetomidine Control RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 95% CI Year

Elective postoperative patients

Venn [8]
Talke [29]
Venn [30]
Triltsch [18]
Martin [20]
Corbett [23]
Wahlander [14]
Maldonado [32]
Shehabi [17]
Subtotal (95% CI)
Total events: 9 (Dexmedetomidine), 13 (Control)
Test for heterogeneity: Chi² = 4.82, df = 7 (P = 0.68), I² = 0%
Test for overall effect: Z = 0.67 (P = 0.50)

Non-elective critically ill patients

Memis [21]
Pandharipande [34]
Memis [26]
Reade [31]
Riker [33]
Ruokonen [25]
Tasdogan [27]
Subtotal (95% CI)
Total events: 80 (Dexmedetomidine), 62 (Control)
Test for heterogeneity: Chi² = 3.10, df = 6 (P = 0.80), I² = 0%
Test for overall effect: Z = 0.96 (P = 0.34)

Total (95% CI)

Total events: 89 (Dexmedetomidine), 75 (Control)
Test for heterogeneity: Chi² = 8.03, df = 14 (P = 0.89), I² = 0%
Test for overall effect: Z = 1.12 (P = 0.26)

0.01 0.1 1 10 100

Favors dexmedetomidine Favors control

Fig. 7 Effect of dexmedetomidine on risk of mortality

to different patient cohorts, ICU discharge criteria, and allocation concealment and double blinding also could
rescue sedative agents between the pooled studies. not confirm this benefit of dexmedetomidine. Second, we
Restricting analysis to studies that had both adequate could not demonstrate a significant reduction in duration
 0.0 Standard error (log RR)

0.4

0.8

1.2

1.6

0.01 0.1 1 10 100

Relative risk
Favors dexmedetomidine Favors alternative
(RR)

Fig. 8 Using hypotension requiring interventions as an endpoint, funnel plot does not suggest the presence of publication bias

of mechanical ventilation and hospital stay by using early mobilization [40], is very useful in improving short-
dexmedetomidine. and long-term outcomes of critically ill patients. It is
Our results showed that dexmedetomidine was asso- possible that the benefits of dexmedetomidine are more
ciated with increased risk of bradycardia requiring significant when used with daily sedation interruption,
interventions when both a loading dose and high main- weaning protocol, and early mobilization. Third, most
tenance doses were used. Nevertheless, the increased risk studies have excluded patients with pre-existing neuro-
of bradycardia was not accompanied by increased risk of logical diseases (58%), delirium, or drug dependency. A
systemic hypotension (Fig. 5) and only required relatively small study comparing dexmedetomidine with haloperi-
simple interventions such as reducing the infusion rate of dol in patients with delirium has indeed shown that
dexmedetomidine in a recent large randomized controlled dexmedetomidine was very effective in reducing duration
study [33]. Many studies have, however, excluded of mechanical ventilation and length of ICU stay [31]. As
patients with pre-existing liver failure (54%), bradycardia such, dexmedetomidine may be more useful, as a sup-
and heart block (46%). As such, the balance between plementary therapy, in some selected subgroups of
benefits and risks of using a loading dose and high critically ill patients.
maintenance doses of dexmedetomidine in these patients This study has some limitations. First, because of
should be carefully considered. significant heterogeneity in the doses of dexmedetomi-
Despite the potential favorable pharmacodynamic dine and characteristics of the patients between the
properties [7, 9], we could not demonstrate a reduction in pooled studies, our results may not be generalizable.
risk of delirium after use of dexmedetomidine. First, Second, many studies used only sedation scale instead of
many patients who were randomized to dexmedetomidine a well-validated method to identify delirium. This could
still required concurrent or rescue alternative sedative or have left patients with hypomania unrecognized.
analgesic agents to achieve the targeted level of sedation Furthermore, this study may also be underpowered to
in the pooled studies. The substantial proportion of demonstrate a small, but significant, benefit on the risk
patients crossed over between two treatment groups of delirium. If the baseline risk of delirium is 40%, a
reduced the power of the studies to demonstrate any sample size of [2,000 patients would be needed to
potential benefits of dexmedetomidine on risk of delirium. demonstrate a 15% relative risk reduction in delirium
Furthermore, many studies also reported the risk of after use of dexmedetomidine. Third, evidence suggests
delirium differently such as ‘‘delirium-free days’’ or that sedation protocols have significant effects on long-
‘‘delirium days’’ and by different cutoffs on different term outcomes of critically ill patients [39]. None of the
sedation assessment scales. Second, although most studies pooled studies have reported on the long-term outcomes
titrated the dose of the study drug according to the level of of their patients and, as such, the effects of using dex-
sedation required (88%), only two studies used mandatory medetomidine on long-term patient-centered outcomes
daily sedation interruption to avoid oversedation [25, 33]. remain uncertain.
Evidence suggests that daily sedation interruption [38], In summary, significant heterogeneity existed between
especially when used with weaning protocol [39] and the pooled studies. The limited evidence suggested that
dexmedetomidine might reduce the length of ICU stay in Acknowledgments The study was solely funded by Department of
some critically ill patients. The risk of bradycardia was, Intensive Care Medicine, Royal Perth Hospital. The funding source
had no role in the collection, analysis, and interpretation of the data
however, higher when both a loading dose and high or in the decision to submit the manuscript for publication.
maintenance doses of dexmedetomidine were used. More
studies are needed to define which subgroups of critically Conflict of interest statement All authors have no financial
ill patients who are most likely to benefit from using interest to declare in relation to the subject matter of this
manuscript.
dexmedetomidine as a primary sedative agent.

References
1. Gravel NR, Searle NR, Sahab PG, 11. Venn RM, Karol MD, Grounds RM 19. Herr DL, Sum-Ping ST, England M
Carrier M (1999) Sedation in critically (2002) Pharmacokinetics of (2003) ICU sedation after coronary
ill patients: practical recommendation. dexmedetomidine infusions for sedation artery bypass graft surgery:
CNS Drugs 11:9–22 of postoperative patients requiring dexmedetomidine-based versus
2. Ho KM, Ng JY (2008) The use of intensive care. Br J Anaesth 88:669– propofol-based sedation regimens. J
propofol for medium and long-term 675 Cardiothorac Vasc Anesth 17:576–584
sedation in critically ill adult patients: 12. Altman DG, Bland JM (2003) 20. Martin E, Ramsay G, Mantz J,
a meta-analysis. Intensive Care Med Interaction revisited: the difference Sum-Ping ST (2003) The role of
34:1969–1979 between two estimates. BMJ 326:219 the alpha2-adrenoceptor agonist
3. Riker RR, Fraser GL (2009) Altering 13. Higgins JP, Thompson SG, Deeks JJ, dexmedetomidine in postsurgical
intensive care sedation paradigms to Altman DG (2003) Measuring sedation in the intensive care unit.
improve patient outcomes. Crit Care inconsistency in meta-analyses. BMJ J Intensive Care Med 18:29–41
Clin 25:527–538 327:557–560 21. Memiş D, Hekimoğlu S, Vatan I,
4. Tan JA, Ho KM (2009) Use of 14. Wahlander S, Frumento RJ, Wagener Yandim T, Yüksel M, Süt N (2007)
remifentanil as a sedative agent in G, Saldana-Ferretti B, Joshi RR, Effects of midazolam and
critically ill adult patients: a meta- Playford HR, Sladen RN (2005) A dexmedetomidine on inflammatory
analysis. Anaesthesia 64:1342–1352 prospective, double-blind, randomized, responses and gastric intramucosal pH
5. Paris A, Tonner PH (2005) placebo-controlled study of to sepsis, in critically ill patients.
Dexmedetomidine in anaesthesia. Curr dexmedetomidine as an adjunct to Br J Anaesth 98:550–552
Opin Anaesthesiol 18:412–418 epidural analgesia after thoracic 22. Memiş D, Dökmeci D, Karamanlioğlu
6. Guo TZ, Jiang JY, Buttermann AE, surgery. J Cardiothorac Vasc Anesth B, Turan A, Türe M (2006) A
Maze M (1996) Dexmedetomidine 19:630–635 comparison of the effect on gastric
injection into the locus ceruleus 15. Akin S, Aribogan A, Arslan G (2008) emptying of propofol or
produces antinociception. Dexmedetomidine as an adjunct to dexmedetomidine in critically ill
Anesthesiology 84:873–881 epidural analgesia after abdominal patients: preliminary study. Eur J
7. Venn RM, Hell J, Grounds RM (2000) surgery in elderly intensive care Anaesthesiol 23:700–704
Respiratory effects of dexmedetomidine patients: a prospective, double-blind, 23. Corbett SM, Rebuck JA, Greene CM,
in the surgical patient requiring clinical trial. Curr Ther Res Clin Exp Callas PW, Neale BW, Healey MA,
intensive care. Crit Care 4:302–308 69:16–28 Leavitt BJ (2005) Dexmedetomidine
8. Venn RM, Bradshaw CJ, Spencer R, 16. Kaneko T (2008) Postoperative does not improve patient satisfaction
Brealey D, Caudwell E, Naughton C, management of carotid endarterectomy when compared with propofol during
Vedio A, Singer M, Feneck R, Treacher with dexmedetomidine—a comparison mechanical ventilation. Crit Care Med
D, Willatts SM, Grounds RM (1999) with propofol. Masui 57:696–703 33:940–945
Preliminary UK experience of 17. Shehabi Y, Grant P, Wolfenden H, 24. Elbaradie S, El Mahalawy FH, Solyman
dexmedetomidine, a novel agent for Hammond N, Bass F, Campbell M, AH (2004) Dexmedetomidine vs.
postoperative sedation in the intensive Chen J (2009) Prevalence of delirium propofol for short-term sedation of
care unit. Anaesthesia 54:1136–1142 with dexmedetomidine compared with postoperative mechanically ventilated
9. Huupponen E, Maksimow A, morphine based therapy after cardiac patients. J Egypt Natl Cancer Inst
Lapinlampi P, Särkelä M, Saastamoinen surgery: a randomized controlled trial 16:153–158
A, Snapir A, Scheinin H, Scheinin M, (DEXmedetomidine COmpared to 25. Ruokonen E, Parviainen I, Jakob SM,
Meriläinen P, Himanen SL, Morphine-DEXCOM Study). Nunes S, Kaukonen M, Shepherd ST,
Jääskeläinen S (2008) Anesthesiology 111:1075–1084 Sarapohja T, Bratty JR, Takala J,
Electroencephalogram spindle activity 18. Triltsch AE, Welte M, von Homeyer P, ‘‘Dexmedetomidine for Continuous
during dexmedetomidine sedation and Grosse J, Genähr A, Moshirzadeh M, Sedation’’ Investigators (2009)
physiological sleep. Acta Anaesthesiol Sidiropoulos A, Konertz W, Kox WJ, Dexmedetomidine versus propofol/
Scand 52:289–294 Spies CD (2002) Bispectral index- midazolam for long-term sedation
10. Sichrovsky TC, Mittal S, Steinberg JS guided sedation with dexmedetomidine during mechanical ventilation.
(2008) Dexmedetomidine sedation in intensive care: a prospective, Intensive Care Med 35:282–290
leading to refractory cardiogenic shock. randomized, double blind, placebo-
Anesth Analg 106:1784–1786 controlled phase II study. Crit Care
Med 30:1007–1014
26. Memiş D, Kargi M, Sut N (2009) 32. Maldonado JR, Wysong A, van der 37. Dasta JF, Kane-Gill SL, Pencina M,
Effects of propofol and Starre PJ, Block T, Miller C, Reitz BA Shehabi Y, Bokesch PM, Wisemandle
dexmedetomidine on indocyanine green (2009) Dexmedetomidine and the W, Riker RR (2010) A cost-
elimination assessed with LIMON to reduction of postoperative delirium minimization analysis of
patients with early septic shock: a pilot after cardiac surgery. Psychosomatics dexmedetomidine compared with
study. J Crit Care 24:603–608 50:206–217 midazolam for long-term sedation in the
27. Tasdogan M, Memis D, Sut N, Yuksel 33. Riker RR, Shehabi Y, Bokesch PM, intensive care unit. Crit Care Med
M (2009) Results of a pilot study on the Ceraso D, Wisemandle W, Koura F, 38:497–503
effects of propofol and Whitten P, Margolis BD, Byrne DW, 38. Kress JP, Pohlman AS, O’Connor MF,
dexmedetomidine on inflammatory Ely EW, Rocha MG, SEDCOM (Safety Hall JB (2000) Daily interruption of
responses and intraabdominal pressure and Efficacy of Dexmedetomidine sedative infusions in critically ill
in severe sepsis. J Clin Anesth 21:394– Compared With Midazolam) Study patients undergoing mechanical
400 Group (2009) Dexmedetomidine vs ventilation. N Engl J Med 342:1471–
28. Talke P, Li J, Jain U, Leung J, Drasner midazolam for sedation of critically ill 1477
K, Hollenberg M, Mangano DT (1995) patients: a randomized trial. JAMA 39. Girard TD, Kress JP, Fuchs BD,
Effects of perioperative 301:489–499 Thomason JW, Schweickert WD, Pun
dexmedetomidine infusion in patients 34. Pandharipande PP, Pun BT, Herr DL, BT, Taichman DB, Dunn JG, Pohlman
undergoing vascular surgery. The study Maze M, Girard TD, Miller RR, AS, Kinniry PA, Jackson JC, Canonico
of perioperative ischemia research Shintani AK, Thompson JL, Jackson AE, Light RW, Shintani AK, Thompson
group. Anesthesiology 82:620–633 JC, Deppen SA, Stiles RA, Dittus RS, JL, Gordon SM, Hall JB, Dittus RS,
29. Talke P, Chen R, Thomas B, Aggarwall Bernard GR, Ely EW (2007) Effect of Bernard GR, Ely EW (2008) Efficacy
A, Gottlieb A, Thorborg P, Heard S, sedation with dexmedetomidine vs and safety of a paired sedation and
Cheung A, Son SL, Kallio A (2000) The lorazepam on acute brain dysfunction in ventilator weaning protocol for
hemodynamic and adrenergic effects of mechanically ventilated patients: the mechanically ventilated patients in
perioperative dexmedetomidine infusion MENDS randomized controlled trial. intensive care (awakening and
after vascular surgery. Anesth Analg JAMA 298:2644–2653 breathing controlled trial): a
90:834–839 35. Ozkan N, Sahin N, Karakaya Kabukcu randomised controlled trial. Lancet
30. Venn RM, Grounds RM (2001) H, Celikbilek G, Aydogdu Titiz T 371:126–134
Comparison between dexmedetomidine (2007) Comparison of the effects of 40. Schweickert WD, Pohlman MC,
and propofol for sedation in the dexmedetomidine with midasolam on Pohlman AS, Nigos C, Pawlik AJ,
intensive care unit: patient and clinician haemodynamics and oxygen profile in Esbrook CL, Spears L, Miller M,
perceptions. Br J Anaesth 87:684–690 patients undergoing coronary artery Franczyk M, Deprizio D, Schmidt GA,
31. Reade MC, O’Sullivan K, Bates S, bypass grafting. Gogus-Kalp-Damar Bowman A, Barr R, McCallister KE,
Goldsmith D, Ainslie WR, Bellomo R Anestezi ve Yogun Bakim Dernegi Hall JB, Kress JP (2009) Early physical
(2009) Dexmedetomidine vs. Dergisi 13:161–168 and occupational therapy in
haloperidol in delirious, agitated, 36. Esmaoglu A, Ulgey A, Akin A, Boyaci mechanically ventilated, critically ill
intubated patients: a randomised open- A (2009) Comparison between patients: a randomised controlled trial.
label trial. Crit Care 13:R75 dexmedetomidine and midazolam for Lancet 373:1874–1882
sedation of eclampsia patients in the
intensive care unit. J Crit Care
24:551–555