Drugs 2012; 72 (14): 1881-1916

REVIEW ARTICLE 0012-6667/12/0014-1881/$55.55/0

Adis ª 2012 Springer International Publishing AG. All rights reserved.

Sedation for Critically Ill or Injured Adults

in the Intensive Care Unit
A Shifting Paradigm
Derek J. Roberts,1 Babar Haroon2 and Richard I. Hall2
1 Departments of Surgery, Community Health Sciences (Division of Epidemiology) and Critical Care
Medicine, University of Calgary and the Foothills Medical Centre, Calgary, AB, Canada
2 Department of Anesthesia and the Division of Critical Care Medicine, Dalhousie University and the
Queen Elizabeth II Health Sciences Centre, Halifax, NS, Canada

Contents
Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1882
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1882
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1883
3. A Shifting Paradigm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1883
4. Indications for Uninterrupted or Prolonged Sedation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1884
5. The Importance of Analgesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1884
6. Basic Pharmacological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1885
6.1 Pharmacodynamic Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1885
6.1.1 GABAA Receptor Agonists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1885
6.1.2 Agonists of the a2-Adrenergic Receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1886
6.2 Pharmacokinetic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1886
6.2.1 Changes in Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1886
6.2.2 Changes in Drug Metabolism, Transport and Elimination . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
7. Tools for Monitoring and Assessing the Adequacy of Sedation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
8. Adverse Consequences of Sedation Common Across Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1888
8.1 Delirium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
8.2 Prolonged Drug Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
8.3 Sleep Disturbance and Withdrawal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
8.4 Immunity and Intensive Care Unit-Acquired Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
9. Specific Drug Classes and Sedative Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1890
9.1 Benzodiazepines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1890
9.2 Anaesthetic Agents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1900
9.2.1 Propofol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1900
9.2.2 Volatile Inhalational Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1901
9.2.3 Etomidate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1901
9.2.4 Ketamine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902
9.3 a2-Adrenergic Receptor Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902
9.3.1 Clonidine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902
9.3.2 Dexmedetomidine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902
10. Pharmacoeconomic Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1903
11. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1905
1882 Roberts et al.

Abstract As most critically ill or injured patients will require some degree of seda-
tion, the goal of this paper was to comprehensively review the literature
associated with use of sedative agents in the intensive care unit (ICU). The
first and selected latter portions of this article present a narrative overview of
the shifting paradigm in ICU sedation practices, indications for uninterrupted
or prolonged ICU sedation, and the pharmacology of sedative agents. In the
second portion, we conducted a structured, although not entirely systematic,
review of the available evidence associated with use of alternative sedative
agents in critically ill or injured adults. Data sources for this review were
derived by searching OVID MEDLINE and PubMed from their first avail-
able date until May 2012 for relevant randomized controlled trials (RCTs),
systematic reviews and/or meta-analyses and economic evaluations.
Advances in the technology of mechanical ventilation have permitted
clinicians to limit the use of sedation among the critically ill through daily
sedative interruptions or other means. These practices have been reported to
result in improved mortality, a decreased length of ICU and hospital stay
and a lower risk of drug-associated delirium. However, in some cases, pro-
longed or uninterrupted sedation may still be indicated, such as when patients
develop intracranial hypertension following traumatic brain injury. The
pharmacokinetics of sedative agents have clinical importance and may be
altered by critical illness or injury, co-morbid conditions and/or drug-drug
interactions. Although use of validated sedation scales to monitor depth of
sedation is likely to reduce adverse events, they have no utility for patients
receiving neuromuscular receptor blocking agents. Depth of sedation mon-
itoring devices such as the Bispectral Index (BISª) also have limitations.
Among existing RCTs, no sedative agent has been reported to improve the
risk of mortality among the critically ill or injured. Moreover, although
propofol may be associated with a shorter time to tracheal extubation and
recovery from sedation than midazolam, the risk of hypertriglyceridaemia
and hypotension is higher with propofol. Despite dexmedetomidine being
linked with a lower risk of drug-associated delirium than alternative sedative
agents, this drug increases risk of bradycardia and hypotension. Among
adults with severe traumatic brain injury, there are insufficient data to suggest
that any single sedative agent decreases the risk of subsequent poor neuro-
logical outcomes or mortality. The lack of examination of confounders,
including the type of healthcare system in which the investigation was
conducted, is a major limitation of existing pharmacoeconomic analyses,
which likely limits generalizability of their results.

1. Introduction systems[5,6] are increasingly utilized in an attempt

to improve the quality of sedation among the
Intensive care unit (ICU) patients requiring me- critically ill,[7] recent data suggest that sedation
chanical ventilatory support commonly receive sed- management varies considerably across ICUs,[8]
ative agents. Indications for ICU sedation may and continues to be suboptimal.[9-11] Moreover,
include reduction of anxiety or agitation, resolu- despite the introduction of evidence-based recom-
tion of ventilator dysynchrony and management mendations on use of sedation and analgesia in
of raised intracranial pressure (ICP), among many critically ill adults,[12,13] very few sedative agents
others.[1,2] Although protocols[3,4] and scoring have been rigorously evaluated by more than one

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1883

or two randomized controlled trials (RCTs),[14] pofol’, ‘dexmedetomidine’, ‘ketamine’, ‘etomidate’

which limits the confidence and strength of ex- and ‘anesthetics’, each combined with the MeSH
isting recommendations. terms ‘critical care’, ‘critical illness’, ‘intensive care’
As the pharmacology of sedative agents in and ‘intensive care unit’. One reviewer (RIH) then
the setting of critical illness or injury is complex, screened the titles and abstracts of all identified
and because our understanding of the role of se- articles and selected relevant RCTs, systematic
dation during mechanical ventilation has recently reviews and/or meta-analyses and economic eval-
evolved,[15-17] the goal of this paper was to com- uations for inclusion in the review. This reviewer
prehensively review the literature associated with also independently extracted data from included
use of sedative agents in the ICU. More specifi- articles in order to construct summary RCT, syste-
cally, our objectives were to review (i) the shifting matic review and cost-effectiveness evidentiary
paradigm in ICU sedation practices; (ii) indica- tables.
tions for prolonged or uninterrupted sedation
among critically ill adults; (iii) the importance of 3. A Shifting Paradigm
adequate analgesia among sedated ICU patients;
(iv) the pharmacology of sedative agents; (v) tools The requirement for sedation among critically
for monitoring and assessing the adequacy of seda- ill patients receiving mechanical ventilation dates
tion; (vi) sedative-related adverse effects; (vii) the back to the beginnings of intensive care medicine.
efficacy of sedative agents in the ICU; and As initial ventilators were relatively crude de-
(viii) economic implications of alternate sedative vices, it was often necessary to employ pharma-
regimes. For the purposes of this review, we di- cological muscle relaxation in order for patients
vided sedative agents into the following drugs or to tolerate existing modes of ventilation. Seda-
drug classes: (i) benzodiazepines (diazepam, lora- tion was therefore indicated to avoid the dis-
zepam and midazolam); (ii) anaesthetic agents comfort linked with mechanical ventilation and
(etomidate, ketamine, propofol and volatile in- neuromuscular blockade.
halational agents); and (iii) a2-adrenergic receptor However, when advances in the technology of
agonists (clonidine and dexmedetomidine). Al- mechanical ventilation allowed patients to be awake
though use of opioid analgesics as a component of and breathing spontaneously, a paradigm shift oc-
analgo-sedation is common in many ICU settings, curred with respect to sedation in critical care
opioids were not part of the intended focus of this medicine. This shift was accelerated by the results of
review and therefore were not specifically con- a single-centre RCT by Kress et al.[18] conducted in
sidered herein. a medical ICU. This study enrolled 128 adults and
demonstrated that daily sedation interruption in
2. Methods mechanically ventilated patients (until they were
awake and able to follow commands) led to a
The first and selected latter portions of this shorter duration of mechanical ventilation and ICU
article present a narrative overview of the shifting stay, as well as a reduction in the number of diag-
paradigm in ICU sedation practices, indications nostic imaging tests for altered mental status.[18]
for prolonged or uninterrupted ICU sedation, Since this study, accumulating evidence has
and the pharmacology of sedative agents. In the suggested that a reduction in the degree of sedation
second portion, we conducted a structured, among critically ill patients may improve clinical
although not entirely systematic, review of the avail- outcomes. In a follow-up nested prospective co-
able evidence supporting use of alternative seda- hort study from the same medical ICU, there was
tive agents in critically ill adults. We searched a trend toward a better total Pyschosocial Ad-
personal files and the OVID MEDLINE and justment to Illness score and a lower incidence
PubMed databases from their first available date of post-traumatic stress disorder (PTSD) among
until May 2012, using the exploded Medical Subject those receiving daily sedation interruption.[19]
Heading (MeSH) terms ‘benzodiazepines’, ‘pro- A second RCT demonstrated that a daily spon-

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1884 Roberts et al.

taneous awakening trial followed by a spontaneous Table I. Indications for uninterrupted or prolonged sedation among
breathing trial improved mortality and decreased critically ill or injured adults[2,29-33]
the length of mechanical ventilation and ICU and ALI/ARDS
hospital stay versus usual sedation care plus a  Or other conditions where the extent of hypoxia and/or
hypercarbia is such that alternative ventilation strategies, use of
daily spontaneous breathing trial.[20] Finally, a
a neuromuscular blocking agent or ECMO is required
fourth RCT suggested that early exercise and
Severe TBI or other aetiologies of significantly reduced intracranial
mobilization through physical and occupational compliance and one or more of:
therapy during periods of sedation interruption  Accumulation of airway secretions (as suctioning may produce
may reduce the risk of delirium and allow for an rapid and large increases in ICP among patients who are not
earlier return to independent functional status.[21] sedated)
Importantly, with one possible exception,[22]  Agitation (which may elevate ICP through increases in
intrathoracic pressure and other mechanisms)
studies have not associated daily awakening or
 Intracranial hypertension (as non-opioid sedation often has
sedative interruptions with an increased risk of favourable effects on ICP and CPP)
adverse events, including self-extubation, among Haemodynamic instability associated with myocardial infarction/
critically ill adults.[18,20,23-25] Even in an RCT that ischaemia
examined the role of no sedation, the no sedation  In order to reduce myocardial oxygen demand and enable
group had fewer days of mechanical ventilation revascularization or institution of intra-aortic balloon
and shorter ICU and hospital lengths of stay,[26] counterpulsation or another form of mechanical haemodynamic
support
without an increased risk of adverse long-term
Intoxications where the agitation of the patient is such that he or she
neuropsychological outcomes.[27] could be a harm to themself or others
Given the beneficial outcomes associated with ALI = acute lung injury; ARDS = acute respiratory distress syndrome;
sedation-minimization outlined above, there seems CPP = cerebral perfusion pressure; ECMO = extracorporeal membrane
little indication to continue a practice of deep se- oxygenation; ICP = intracranial pressure; TBI = traumatic brain injury.

dation for mechanically ventilated patients in the

absence of a requirement for a neuromuscular re- Although amnesia is often cited as an indication
ceptor blocking agent. However, to do so requires for sedation during mechanical ventilation,[36] the
careful attention to the level of sedation provided, credence of this assumption is increasingly being
which appears to be best guided by use of proto- challenged. Complete amnesia for prolonged peri-
colized, target-based sedation regimes,[28] and an ods of time during mechanical ventilation in the
in-depth understanding of the pharmacology of ICU has not been shown to provide benefit, and
ICU sedation. could potentially cause harm. In support of this,
mounting evidence suggests that critically ill pa-
4. Indications for Uninterrupted or tients may benefit from retaining an awareness of
Prolonged Sedation their surroundings, as this may result in improved
long-term neuropsychiatric recovery and clinical
Although uninterrupted or prolonged sedation outcomes following critical illness.[19,37,38]
should generally be avoided in critically ill patients,
in some scenarios it may be indicated (table I).[2,29-33] 5. The Importance of Analgesia
However, in these cases, once the primary dis-
turbance has been adequately controlled or treat- Sedation may mask pain,[39] and pain is a risk
ed, sedation should be reduced to the minimal factor for development of agitation when in-
level required and attempts made to remove the appropriately managed.[40,41] Sources of pain and
sedative at least daily.[18,20] Absent the use of neuro- discomfort among critically ill or injured patients
muscular receptor blocking agents, even patients receiving mechanical ventilation include manip-
being treated with low tidal volumes as part of a ulation of post-traumatic injuries such as fractured
ventilation strategy for management of acute ribs and extremities, wound care and dressing
respiratory distress syndrome (ARDS) may be changes, performance of invasive procedures (in-
managed with minimal sedation.[34,35] cluding insertion of vascular monitors or chest

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1885

drains), routine nursing care (including patient re- agents as well as specific sedative adverse drug
positioning), suctioning of airway secretions via the reactions.[2,12,44-52]
endotracheal tube and prolonged immobility with
resultant joint stiffness.[42,43] Therefore, clinicians 6.1 Pharmacodynamic Effects
should ensure that adequate analgesia is achieved
The most common sedative agents employed
before starting any sedative agent or sedation regime.
for the provision of sedation in the ICU act via
the gamma-amino-butyric-acid type A (GABAA)
6. Basic Pharmacological Considerations receptor. Two agonists of the a2-adrenergic re-
ceptor also exist.
A comparison of the pharmacodynamics, phar-
macokinetics and important drug interactions of 6.1.1 GABAA Receptor Agonists
sedative agents is shown in table II.[12,44-49] Table The GABAA receptor is a transmembrane chlo-
III affords an overview of the effects of organ ride channel positioned on neurons throughout the
dysfunction on the clearance of individual sedative central nervous system (CNS). Binding of sedative

Table II. Pharmacodynamics, pharmacokinetics and important drug interactions of sedative agentsa[12,44-49]
Sedative agent Onset after Terminal Metabolism Active metabolites Important drug interactions with
IV bolus t½ (h) substrates/inhibitors ( › effect) and
(min) inducers ( fl effect) of CYPb
Benzodiazepines (GABAA receptor agonists)
Diazepam 2–5 21–120 CYP3A4 and 2C19 Desmethyldiazepam and CYP3A4 substrates/inhibitors
oxazepam ( › effect) or inducers ( fl effect)
and CYP2C19 substrates/inhibitors
(e.g. cimetidine, omeprazole)
Lorazepam 5–20 10–14 Glucuronidation None Valproate ( › effect)
Midazolam 2–5 7–11 CYP3A4 a-hydroxymidazolam and CYP3A4 substrates/inhibitors
a-hydroxymidazolam ( › effect) or inducers ( fl effect)
glucuronide
Anaesthetic agents (predominant mechanism of action)
Propofol (GABAA 1–2 26–32 Glucuronidation and None None
receptor agonist) sulfation
Volatile (complex <1 Varies Predominantly pulmonary None None
pharmacodynamics) clearance
Etomidatec (GABAA ~1 3–6 Hepatic esterases None None
receptor agonist)
Ketamine (NMDA ~1 2 CYP2B6, 2C8/9 and 3A4 Norketamine CYP2B6, 2C8/9 and CYP3A4
receptor antagonist) substrates/inhibitors ( › effect) and
inducers ( fl effect)
a2-Adrenergic receptor agonists
Clonidine NA 6–20 Hepatic None None
Dexmedetomidine 15 2 CYP2A6 and None CYP2A6 substrates/inhibitors
glucuronidation ( › effect) and inducers ( fl effect)
a Whenever possible, data were derived from critically ill adults, and ranges represent the 95% confidence interval.
b Common CYP3A4 substrates include codeine, diazepam, fentanyl, haloperidol, ketamine and midazolam, while inhibitors include
cimetidine, clarithromycin, diltiazem, erythromycin, azole antifungal agents and verapamil; inducers include barbiturates, carbamazepine,
phenytoin and rifampicin (rifampin).[44]
c Only approved for use as an induction agent for rapid sequence intubation among critically ill or injured patients and not for continuous IV
infusion (see text for explanation).
CYP = cytochrome P450; GABAA = gamma-amino-butyric-acid type A; IV = intravenous; NA = not applicable; NMDA = N-methyl-D-aspartate
glutamatergic receptor; t½ = half-life; › indicates increased; fl indicates decreased.

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1886 Roberts et al.

Table III. Specific sedative adverse drug reactions and effects of organ dysfunction on sedative action[2,12,44-51]
Sedative agent Effect of organ dysfunction on drug elimination Specific adverse drug reactions
Renal dysfunction End-stage liver disease
Benzodiazepines
Diazepam No effect to mild effect Major effect Hypotension, respiratory depression, paradoxical agitation, withdrawal
Lorazepam No effect Major effect syndrome with acute discontinuation, precipitation in intravenous lines
(lorazepam) and propylene glycol toxicity (lorazepam)
Midazolam Major effect Major effect
Anaesthetic agents
Propofol No effect No effect Injection site pain, respiratory depression and apnoea, mild
myocardial depression, green discoloration of the urine, hypotension,
hyperlipidaemia, pancreatitis and propofol infusion syndrome
(metabolic acidosis, hyperkalaemia, rhabdomyolysis,
hyperlipidaemia and/or an enlarged fatty liver)
Volatile No effect No effect Arrythmia and malignant hyperthermia
Etomidate No effect Moderate effect Injection site pain, myoclonus, adrenal suppression and increased
mortality following prolonged infusion (and possibly after a single
bolus dose)
Ketamine No effect No effect Psychiatric and emergence effects (hallucinations during dissociative
anaesthesia, emergence delirium and out-of-body experiences),
hypersalivation, lacrimation and increased myocardial oxygen
demand (although originally associated with increased intracranial
pressure, this observation has not been consistently made in
internally valid clinical trials)
a2-Adrenergic receptor agonists
Clonidine Mild effect Insufficient data Xerostomia, bradycardia, hypotension, rebound hypertension and
tachycardia following acute discontinuation
Dexmedetomidine No effect Major effect Xerostomia, bradycardia and hypotension

drugs to this channel increases either the frequency more organ systems and exhibit significant inter-
or duration of channel opening, which facilitates individual variation in drug response.[58-61] More-
neuronal membrane hyperpolarization (by pro- over, this patient population often has pre-existing
moting transmembrane chloride exchange) and organ dysfunction that may affect drug disposition
inhibition of subsequent neurotransmission.[53] Se- and metabolism. This may occur as a result of ad-
dative agents thought to produce sedation by vanced age,[62,63] other co-morbid conditions[64,65]
this mechanism include propofol,[54] etomidate, or genetic predisposition.[45,66] Collectively, the
the benzodiazepines,[55] barbiturates[53] and per- available pharmacokinetic evidence suggests that
haps the volatile anaesthetic agents.[50] there is a substantial risk for ICU patients receiving
sedative agents to be either over- or under-sedated.
6.1.2 Agonists of the a2-Adrenergic Receptor
The a2-adrenergic receptor is a pre-synaptic re- 6.2.1 Changes in Distribution
ceptor involved in the release of norepinephrine
As most sedative drugs are lipophilic, and the
(noradrenalin) and modulation of post-synaptic
CNS receives approximately 20% of the cardiac
GABA receptor activation.[56] The selective agonist
output, sedative drugs rapidly distribute into the
dexmedetomidine and the related non-specific ago-
highly perfused, lipid-dense brain following in-
nist clonidine act by stimulation of this receptor.[57]
travenous administration. Subsequently, these
6.2 Pharmacokinetic Considerations
drugs re-distribute out of the brain and into less
well perfused peripheral tissues (e.g. adipose tis-
A thorough understanding of the pharmaco- sue and skeletal muscle) after one or multiple
kinetics of sedative agents is required as critically ill bolus doses, resulting in a rapid offset of phar-
patients frequently display dysfunction of one or macodynamic effect. However, once peripheral

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1887

tissues are saturated (often following prolonged lowing cardiac arrest or severe traumatic brain
drug infusion), redistribution ceases to occur, and injury has also been reported to alter sedative
the offset of drug effect becomes dependent on agent pharmacokinetics substantially.[76,77]
the elimination half-life of the drug.[67] This con- In addition to drug elimination, drug transport
cept (that the elimination of a drug is dependent on may be altered in critically ill or injured adults.[78]
the duration of administration) is important in the Central and peripheral inflammatory responses to
practice of critical care medicine, and encompassed injury, surgery and/or sepsis may down- or upreg-
by the term ‘context-sensitive half-time’.[68] ulate various drug transporters at the level of
the gut, kidney or blood-brain barrier. Emerging
6.2.2 Changes in Drug Metabolism, Transport evidence in animals[78] suggests that the expression
and Elimination and function of the P-glycoprotein multi-drug trans-
Sedative drug metabolism and elimination may porter may be transiently down- or up-regulated by
be altered by several factors in critically ill or in- inflammation, which could potentially increase
jured patients.[69,70] Excretion of active metabo- or decrease CNS drug effects. In support of this,
lites may be impaired among those with impaired the cerebrospinal fluid (CSF) concentration of
renal clearance following acute kidney injury or morphine’s glucuronidated metabolites increased
other insults.[71] Competitive inhibition of cyto- linearly with the CSF concentration of the pro-
chrome P450 (CYP) drug metabolizing enzymes inflammatory cytokine interleukin-6 following
may also occur when two drugs are metabolized acute brain injury in a recent prospective phar-
by the same isoenzyme. This mode of inhibition macokinetic study.[79]
frequently affects midazolam metabolism and
clearance as this agent is metabolized by the en- 7. Tools for Monitoring and Assessing the
zyme CYP3A, which also metabolizes other drugs Adequacy of Sedation
commonly used among the critically ill, including
diltiazem, fentanyl, ketamine and propofol.[72] As uninterrupted or prolonged sedation is asso-
Thus, co-administration of propofol may com- ciated with worsened outcomes for most critically
petitively inhibit midazolam metabolism, increase ill patients,[80] there are compelling reasons to moni-
drug concentrations and result in a prolonged tor the level of sedation and adjust it according to
sedative effect.[73,74] In contrast, when midazolam pre-defined treatment objectives. Over-sedation
is co-administered with rifampicin (rifampin) [a may be unrecognized by caregivers,[81] and a proto-
known CYP3A-inducing agent], enhanced mid- colized approach to monitoring and adjusting
azolam metabolism and clearance may lead to sedation levels may lead to improved outcomes.[28]
decreased drug concentrations and a reduction in However, a recent survey conducted in Canadian
the degree and/or duration of sedative effect.[75] ICUs suggested that only a minority utilize seda-
A number of organ and body system changes tion protocols.[82] Moreover, even in ICUs where
occur during the development of critical illness sedation protocols exist, observational studies in-
that may also result in altered pharmacokine- dicate poor compliance with their use, suggesting
tics.[46] These include hepatic dysfunction (which the need for ongoing quality improvement and
may result in reduced CYP enzyme function, portal knowledge translation.[8]
venous and/or hepatic arterial blood flow and An individualized approach to the appropriate
thus drug delivery to the liver, and bile flow and level of sedation should be sought among ICU
possibly biliary drug excretion, as well as altered patients when sedative therapy is indicated. In
albumin production and drug protein binding), some cases, deeper levels of sedation will be war-
gastrointestinal dysfunction and malnutrition, ranted (e.g. management of intracranial hyperten-
and renal failure (with resultant impaired drug sion), while at other times (e.g. in the weaning
excretion as well as changes in total body water, phase of liberation from mechanical ventilatory
albumin concentration and the volume of dis- support) a lighter level of sedation will be appro-
tribution of drugs).[46] Use of hypothermia fol- priate. To avoid excessive sedation, some monitor

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1888 Roberts et al.

of drug effect, such as a sedation scale, will help significant cause of misinterpretation of the BISª
ensure that infusion rates are adjusted according to score, making its value for monitoring of sedation
already pre-defined clinical endpoints.[83] in patients not receiving neuromuscular blockade
There are several sedation scales that are pre- uncertain.[113-117] Moreover, the BISª score may
sently in use in ICU practice. However, few have vary with different versions of the software[118] and
been adequately validated[84] or compared in with age, even at the same level of sedation.[108] In
large prospective trials.[85] Although the Ramsay one study, 83.9% of patients identified as being
Sedation Scale is the oldest,[86] the Sedation Agi- heavily sedated using the Adaptation to Intensive
tation Scale (SAS)[87] and Richmond Agitation- Care Environment (ATICE) score had BISª values
Sedation Scale (RASS) have been tested more that were surprisingly in excess of 60 (a value be-
extensively for validity and reliability.[83,85,88] Un- lieved to correlate with awareness) at least once.[119]
fortunately, all of these scales require some inter- In another study, there was a poor correlation be-
action (verbal or motor) with the patient, and thus tween the BISª score and plasma concentrations of
they are invalid and should not be used for patients lorazepam.[112] This finding was also reported by a
receiving neuromuscular receptor blocking agents, 2006 overview of 19 studies comparing the BISª
as a possibility exists that these patients could still with other sedation scales, which identified poor
be awake and paralyzed.[89] correlation between sedative drug doses and BISª
In order to prevent this possibility from occur- scores, and recommended that ‘‘further studies be
ring, a number of monitors have been introduced conducted to determine the optimal method of
that use processed electroencephalographic (EEG) obtaining BIS data and to evaluate the impact of
activity to assess neurological function in sedated BIS on relevant patient outcomes.’’[120]
patients.[90] The most extensively studied is the Thus, although there have been improvements,
Bispectral Index (BISª, Aspect Medical Systems, based on our review of a representative sampling of
Norwood, MA, USA). The BISª uses a proprie- the more recent literature,[97,99,100,102-112,115-117,119,121-122]
tary algorithm, which takes the EEG signal from the reported concerns regarding BISª still remain
a montage of leads placed on the forehead of the valid. We suggest that further appropriately de-
patient and then processes it and produces a signed studies be conducted before BISª monitor-
number reflective of brain activity ranging from ing be routinely adopted in the ICU. In our opinion,
0 to 100.[91,92] its current value (and perhaps that of other similar
Although the value of the BISª as a monitor of [but less well investigated] brain function moni-
depth of sedation is disputed,[93] patients are gen- tors, including the Narcotrendª [Monitortechnik
erally thought to be anaesthetized at BIS scores GMBH, Bad Bramstedt, Germany],[123] Patient
ranging from 40 to 60. Therefore, this range may State Index [PSArray-2ª – Physiometrix Inc.,
be representative of an appropriate depth of seda- Billerica, MA, USA],[106] Entropyª Module [GE
tion among mechanically ventilated ICU patients Healthcare, Wauwatosa, WI, USA] [spectral en-
receiving neuromuscular receptor blocking agents. tropy][121] and SNAP IIª [Stryker, Kalamazoo,
In support of this, scores above 60 increase the risk MI, USA][122]) may lie in the ability to provide
that the patient may be aware of their surround- some indication of the degree of sedation re-
ings,[94,95] while scores below 40 for greater than quired among patients receiving neuromuscular
5 minutes have been associated with mortality and receptor blocking agents.
adverse long-term cardiovascular outcomes.[96]
Several studies have compared the BISª with a 8. Adverse Consequences of Sedation
variety of sedation scales, including the RASS,[97-100] Common Across Agents
Ramsay Sedation Scale,[101-109] Glasgow Coma
Scale (GCS)[98,110] and the SAS,[98,111,112] for asses- Although each sedative is associated with
sing adequacy of sedation. However, absent the specific adverse effects (see below and table III),
presence of neuromuscular blockade, the presence sedation has several adverse consequences that
of electromyographic activity (e.g. shivering) is a are common across agents.[124]

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1889

8.1 Delirium rapid eye movement [REM] and non-rapid eye

movement [NREM]).[140,141] Sleep disturbance is
Sedative use is consistently associated with the a common finding in ICU patients and may be
development of delirium,[125,126] defined as a dis- exacerbated by a number of factors, including
turbance of consciousness characterized by an pre-morbid illness,[142] pain, diagnostic testing,
acute onset and fluctuating course of impaired noise, and care-related activities, such as tracheal
cognitive functioning such that the patient’s ability suctioning and monitoring of vital signs and
to receive, process, store and recall information is neurological status.[143-146] This problem is ex-
impaired.[127] While lorazepam and other benzo- acerbated among critically ill patients requiring
diazepines may be associated with the greatest risk mechanical ventilatory support.[147]
of drug-associated delirium, dexmedetomidine ap- Although sedatives are commonly given to
pears to be associated with a low risk for this promote sleep in ICU patients, there are emerging
condition. As delirium is associated with increased data that the type of sleep induced by these agents
mortality,[128,129] prolonged duration of mechan- may differ in important aspects from normal
ical ventilation[129] and hospitalization,[128,129] as sleep.[141] Most importantly, although the effect
well as heightened costs[130] and an increased risk on sleep patterns may vary depending on the
of cognitive impairment[131] among the critically ill, mechanism of sedative action,[141,148] there is a
strategies that attempt to limit sedation and there- loss of transition through the sleep cycles, which is
fore avoid delirium may improve outcomes. an effect similar to that observed with sleep dep-
rivation.[141,147,148] This finding is of concern as
8.2 Prolonged Drug Effect sleep deprivation has several associated adverse
Continuous or prolonged administration of sed- consequences, including alterations in immune
ative agents frequently leads to drug accumula- function,[149,150] reductions in respiratory muscle
tion and a prolonged time to being awake and function[151] and even increased mortality.[152]
aware following discontinuation of sedation.[80] Another facet of the interaction between sleep
The longer an infusion continues, the longer it will and sedative use is the possibility of sleep dis-
take for the drug effect to dissipate. This makes the turbance when sedation is stopped. When sleep is
term ‘short-acting’ context sensitive, as the dura- restored after being absent, it frequently takes on
tion of effect will vary depending on the duration of the characteristics of the type of sleep for which
drug infusion. In addition, commonly suggested there has been a deficit (e.g. REM sleep).[141]
clinical dosing guidelines may become inaccurate as REM sleep is the stage of sleep most associated
the duration of sedative infusion increases. Active with cardiovascular changes, and the withdrawal
metabolites,[71,132,133] co-administered drug ther- of REM-suppressive medications (such as ben-
apy[66,134-137] and variations in CYP function may zodiazepines[153]) may lead to a higher incidence
also lead to variations in sedative drug concentra- of haemodynamic perturbations. Although these
tions and prolonged drug effect for the reasons types of withdrawal reactions are common in the
described above. This prolonged drug effect may ICU, they are often unrecognized, and could in-
lead to complications such as delayed liberation fluence patient-important outcomes.[154]
from mechanical ventilation, development of ven-
tilator-associated pneumonia,[138] gastrointestinal 8.4 Immunity and Intensive Care Unit-
dysfunction and aspiration pneumonia[139] and Acquired Infections
drug-associated delirium.[125]
Sepsis-induced organ dysfunction is a leading
8.3 Sleep Disturbance and Withdrawal reason for ICU admission,[155,156] and sepsis is
one of the principal causes of mortality among
Sleep is an essential physiological process, which patients admitted to the ICU.[157] Emerging clin-
in health is characterized by transitions between ical data[158-161] suggest that sedative agents may
stages of the sleep cycle (e.g. slow wave sleep [SWS], increase the risk of developing ICU-acquired

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1890 Roberts et al.

infections. Animal studies (reviewed in Sanders 9.1 Benzodiazepines

et al.[162]) demonstrate that sedative agents influence
both the innate (characterized by host defenses such Benzodiazepines (diazepam, lorazepam and
as barrier function, cytokines, complement, phago- midazolam) are among the most frequently uti-
cytes, natural killer cells and subsets of T cells) and lized sedative agents by practicing intensivists in
adaptive (acquired immunity with a humoral com- North America.[82] These agents act by increasing
ponent involving B lymphocytes and a cellular the frequency of GABA-induced transmembrane
component mediated by T lymphocytes) immune chloride channel opening and have dose-dependent
systems.[162] Although there is insufficient data in anxiolytic, sedative and hypnotic effects.[44,45] They
humans to allow one to differentiate which seda- also raise the seizure threshold and induce ante-
tive agents are the least detrimental, theoretical rograde amnesia and muscle relaxation.[44,45,254]
and experimental data suggest that dexmede- Benzodiazepines have neutral and likely equiva-
tomidine may have a more favourable immuno- lent effects on ICP and CPP among patients with
logical profile.[160,162-164] severe traumatic brain injury compared with the
anaesthetic agent propofol.[2,255] They also may
9. Specific Drug Classes and reduce opioid requirements by modulating the
Sedative Agents development of opioid tolerance or the antici-
patory response to painful stimuli.[256]
In the following section, the pharmacology and Clinically, the benzodiazepines differ mostly
specific adverse drug effects of sedative agents are in their pharmacokinetic profile (table II). Mid-
summarized. The clinical safety and efficacy of azolam has a dissociation constant (pKa) of 6 at
these agents as reported by RCTs and systematic 24C and the solubility of the drug increases as its
reviews and/or meta-analyses are described in solvent pH is reduced.[257] By formulating the
evidentiary tables IV[17,23,26,159,163,165-167,169-250] compound as the hydrochloride salt, midazolam
and V,[2,14,251-253] respectively. Based on the col- hydrochloride becomes water soluble,[258,259] which
lective results of these studies (most of which allows it to have a more rapid onset of action than
examined use of propofol and/or midazolam), the other commonly used benzodiazepines. The
no sedative agent appears to improve the risk hydrophilic nature of the parent compound allows
of mortality among the critically ill or injured. it to be administered intravenously as a bolus or
Moreover, although propofol may be associated continuous infusion with a relatively rapid onset of
with a shorter time to tracheal extubation and action (2–5 minutes),[260] making it especially useful
recovery from sedation than midazolam, the risk for managing acute agitation among critically ill
of hypertriglyceridaemia and hypotension is patients.[12] Offset of action following a single in-
higher with propofol. Despite dexmedetomidine travenous dose also occurs relatively quickly.[44] In
being linked with a lower risk of drug-associated contrast, lorazepam is less useful for the control of
delirium than alternative sedative agents, this agitation as its onset (5–20 minutes) and duration
drug increases the risk of bradycardia and hypo- of action are significantly more protracted.[44,261]
tension. Among adult ICU patients with severe Although diazepam and midazolam undergo
traumatic brain injury, there are insufficient data hepatic and extra-hepatic metabolism by the CYP
from RCTs to suggest that any single sedative system, lorazepam undergoes direct phase II glu-
agent decreases the risk of subsequent poor curonidation in the liver, and is therefore less
neurological outcomes or mortality. Finally, al- susceptible to drug-drug interactions and the ef-
though the clinical significance of this finding is fects of hepatic dysfunction.[44] Diazepam is me-
unknown, several RCTs and a recent systematic tabolized by CYP3A4 and CYP2C19 to the active
review[2] suggest that high bolus doses of opioids metabolites desmethyldiazepam and oxazepam,
may transiently elevate ICP and decrease cerebral while midazolam is converted by CYP3A4 to
perfusion pressure (CPP) among those with se- a-hydroxymidazolam (63% of the potency of
vere head injury. midazolam) and a-hydroxymidazolam glucuronide

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1891

Table IV. Randomized controlled trials comparing alternate sedative agents in critically ill or injured adults
Study Design; Population Main findings Secondary findings
comparators (n)
Jakob et al.,[165] mc, db; DEX (500) Pts requiring prolonged Median duration of MV shorter with Pts treated with DEX had more
2012 vs MID (251) or PRO (>24 h) sedation for DEX vs MID but not vs PRO. DEX hypotension and bradycardia
(247) tolerance of MV pts had improved ability to interact than those receiving MID
with caregivers than with MID or
PRO. ND in ICU or hospital LOS or
mortality
Bjelland et al.,[166] Two-centre, ol; PRO Pts undergoing Offset of sedation significantly ND in pneumonia risk or
2012 and REM (29) vs therapeutic hypothermia shorter and requirements for neurological outcome
MID and FEN (30) following cardiac arrest norepinephrine higher in REM/PRO
and receiving MV group
Strøm et al.,[26] sc, ol; no sedation Adults requiring MV No sedation group had more days Delirium more frequent in the
2010 (70) vs sedation without MV and shorter ICU and no sedation group. ND in
[PRO and MID] (70) hospital LOS incidence of accidental
extubations, VAP or need for
radiological investigations
Hellstrom et al., sc, ol; PRO (50) vs Cardiac surgical pts SEV group had shorter time to ND in ICU or hospital LOS
2011,[167] 2012[168] SEV (49) requiring MV for >2 h extubation and time to adequate
verbal response
Candiotti et al.,[169] mc, ol; FOS Pts requiring MV for FOS groups combined for analysis. FOS-treated pts had more tx-
2011 (bolus/infusion; 18) 2–12 h ND between groups in terms of time associated adverse events
vs FOS (infusion; at targeted sedation level (74% vs 64%), including
20) vs PRO (22) procedural pain and nausea
Mesnil et al.,[170] sc, ol; SEV (19) vs Pts requiring MV for SEV-treated group had shorter Lower incidence of
2011 PRO (14) vs MID >24 h emergence time and time to hallucinations in SEV-treated
(14) extubation. ND in time at target group
sedation level
Mirski et al.,[171] sc, db, crossover; Brain-injured (18) and Degree of return of cognitive More bradycardia in DEX-
2010 DEX vs PRO (30 pts non-brain-injured (12) function improved in the DEX group treated group
total) pts requiring MV among those without brain-injury
(with ND in the brain-injured group)
Tasdogan sc, ol; DEX (20) vs Pts with sepsis requiring Greater reduction in level of intra-
et al.,[172] 2009 PRO (20) MV after abdominal abdominal pressure and
surgery inflammatory cytokines in DEX-
treated group. ND in mortality,
duration of MV or ICU LOS
Esmaoglu sc, ol; DEX (20) vs Women with eclampsia DEX produced differences in initial
et al.,[173] 2009 MID (20) following delivery haemodynamic control in the first
24 h with fewer requirements for
additional antihypertensive agents.
DEX group had a shorter ICU LOS
Riker et al.,[159] mc, db; DEX (244) Medical/surgical pts ND in time spent at target sedation Increased incidence of
2009 vs MID (122) receiving MV for >24 h level. DEX-treated pts had a lower bradycardia in DEX group and
incidence of delirium and a shorter a lower incidence of
time to extubation and ICU LOS hypertension and tachycardia
requiring tx
Ruokonen mc, db, dd; DEX (41) Medical/surgical pts ND in time at target sedation level, Higher incidence of delirium in
et al.,[174] 2009 vs standard care requiring MV for >24 h length of ICU stay or ease of DEX group
[MID or PRO] (44) maintaining a light degree of
sedation. More difficulty acquiring
and maintaining a deep level of
sedation in the DEX group

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1892 Roberts et al.

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Shehabi et al.,[175] mc, db; DEX (162) Pts undergoing ND in time at target sedation level. DEX pts had earlier tracheal
2009 vs morphine (147) cardiac surgery and ND in incidence of delirium but extubation, less hypotension
requiring MV reduced duration of delirium in the and requirement for vasoactive
DEX group agents and an increased
incidence of bradycardia
Maldonado sc, ol; DEX (30) vs Cardiac surgical pts Incidence of delirium significantly Reduced costs in DEX group
et al.,[176] 2009 PRO (30) vs requiring MV reduced in the DEX group
MID (30)
Sackey et al.,[177] sc; ISO (20) vs Medical/surgical pts ND in short-term (4 d)
2008 MID (20) requiring MV >12 h neuropsychological outcome. At
6 mo (in survivors), there was a
trend (p = 0.06) toward improved
neuropsychological outcome in the
ISO vs MID tx group
Huey-Ling sc, ol; PRO (32) vs Postoperative CABG pts ND in haemodynamic changes or
et al.,[178] 2008 MID (28) quality of sedation
Pandharipande Two-centre, db; Medical/surgical pts ND in mortality or cost of care. DEX group had higher
et al.,[179] 2007 DEX (52) vs requiring MV >24 h DEX-treated pts spent more time at incidence of bradycardia (9/52
LOR (51) target sedation level, had more vs 2/51) and self-extubation
days alive without delirium or coma (4/52 vs 2/51)
and a lower prevalence of coma
Ghori et al.,[180] sc, ol; PRO (13) vs Pts with severe head ND in neurological outcome Serum S100b levels were
2007 MID (15) injury requiring sedation higher in pts with poorer
and MV neurological outcome, but did
not differ between groups
Carson et al.,[23] mc, ol; LOR (64) vs Medical pts requiring MV Median ventilator days were lower ND in hospital mortality
2006 PRO (68) for 24 h in PRO group
Richman et al.,[181] sc, ol; MID (13) vs Pts requiring MV for The MID and FEN group spent ND in costs between tx groups
2006 MID and FEN (17) >48 h more time at target sedation level
and had fewer episodes of pt-
ventilator dysynchrony
Hsiao et al.,[182] sc, ol; PRO (51) vs Postoperative pts Amnesia more prevalent in MID
2006 MID (51) following major surgery group. ND in degree of suppression
who required ICU of anxiety
monitoring
Ibrahim et al.,[183] sc, db, pc; melatonin Tracheostomized pts not ND in sleep duration vs PL
2006 (14) vs PL (18) receiving any other
sedation and who were
weaning from MV
Breen et al.,[184] mc, ol; REM (57) vs ICU pts requiring MV for Significant reduction in duration of ND in mortality or serious
2005 MID (48) 3–10 d MV and weaning period observed adverse events
in REM group. ND in time at target
sedation level or pain intensity
Corbett et al.,[185] sc; DEX (43) vs Postoperative elective ND in duration of MV or level of More hypotension in DEX-
2005 PRO (46) CABG pts with duration awareness. Deeper sedation level treated pts
of MV <24 h in PRO group. More reported
discomfort and pain in DEX group
Bourgoin et al.,[186] sc, ol; KET (15) vs Trauma pts with brain 2-fold increases in plasma drug
2005 sufentanil (15) injury requiring MV concentrations did not alter
cerebral haemodynamics

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1893

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Karabinis et al.,[187] sc, ol; REM (84) vs Pts with acute brain Neurological assessment less ND in haemodynamic stability.
2004 PRO/MID and FEN injury or following variable in REM group. Earlier Raters were three times more
(37) or morphine neurosurgery tracheal extubation in REM group likely to choose REM-based
(40) regimen as ‘good’ or ‘excellent’
than other regimens

Martin et al.,[188] mc, pc; DEX (203) Postoperative pts Less supplemental sedation and Hypotension and bradycardia
2003 vs saline (198) requiring MV >6 h morphine required in DEX-treated was more frequent in DEX
group group, while hypertension,
rigors and atelectasis were
more common in control group

Herr et al.,[189] 2003 mc; DEX (148) vs Pts undergoing CABG ND in duration at target sedation Hypertension was more
PRO (147) surgery and requiring level, weaning interval or time to common in DEX group, while
>6 h MV tracheal extubation. Morphine the incidence of ventricular
requirements reduced in DEX tachycardia was higher in PRO
group. Requirements for b- group
adrenergic receptor blocking
agents, antiemetics, NSAIDs,
epinephrine and diuretics lower in
DEX group

Saito et al.,[190] sc, ol; MID (13) vs Surgical pts following Sequential use of MID/PRO
2003 MID and PRO head and neck or resulted in shorter time to tracheal
[sequential] (13) oesophageal surgery extubation and less emergence
requiring MV agitation

Bourgoin et al.,[191] sc, db; KET (12) vs TBI pts requiring MV ND in intra-cranial haemodynamics Higher heart rate in KET vs
2003 sufentanil (13) sufentanil group. Similar
sedation costs

Triltsch et al.,[192] sc, db, pc; DEX (15) Surgical pts requiring Pts receiving DEX required less Morphine requirement reduced
2002 vs PL (15) >6 h MV postoperatively PRO to maintain targeted BISª and in DEX group. Better
during the weaning phase haemodynamic stability in DEX
group

Walder et al.,[193] sc, db, pc; PRO and Male pts following CABG PRO and MID group spent more ND in haemodynamic
2002 PL (26) vs PRO and surgery time at target sedation level. Weaning parameters, blood loss or
MID (27) time prolonged in MID group transfusion requirements

Venn et al.,[194] sc, ol; DEX (10) vs Surgical pts requiring NSD in cortisol, ACTH or glucose Heart rate lower in DEX group.
2001 PRO (10) MV for >8 h concentrations. Higher levels of NSD in arterial pressure.
growth hormone in DEX group. IL-6 Positive response to ACTH
levels decreased in DEX group stimulation test in DEX group
(not tested in PRO group)

Venn and sc, ol; DEX (10) vs Surgical pts requiring ND in target sedation level but PRO Heart rate lower in DEX group
Grounds,[163] 2001 PRO (10) MV for >8 h group received more alfentanil. ND
in time to tracheal extubation

Barr et al.,[195] 2001 sc, db; LOR (14) vs Surgical pts requiring Amnestic properties of LOR 4-
MID (10) MV for up to 72 h times those of MID. More difficult to
achieve stable level of sedation
with LOR use. Prolonged
emergence time and time to
extubation in LOR group

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1894 Roberts et al.

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Hall et al.,[196] 2001 mc, ol; PRO (77) vs Medical/surgical pts PRO-treated pts spent more time at ND in mortality. One death felt
MID (79) requiring MV and target sedation level and had earlier to be directly related to PRO
stratified by short tracheal extubation, but were not administration. Clinically
(<24 h), medium (>24 to discharged from ICU earlier important hypotension (n = 2)
<72 h) and long (>72 h) and hypertriglyceridaemia
duration of MV (n = 1) developed in PRO-
treated pts. Paradoxical
excitement more common in
MID-treated group
Kress et al.,[197] sc, ol (half to daily Medical ICU pts Daily interruption of sedation not ND in ICU LOS or neurological
2001 awakening); PRO requiring MV associated with any increase in examinations between MID or
(62) vs MID (66) adverse events PRO groups
Herr et al.,[198] 2000 mc, db; PRO (63) vs Surgical/trauma pts Lower mortality in PRO EDTA ND in incidence of hypotension,
PRO EDTA (59) requiring MV for >2 h group. ND in quality of sedation at atrial fibrillation,
target sedation levels hypomagnesaemia or
hypocalcaemia. More pts
withdrawn for safety reasons in
PRO group
Barr et al.,[199] 2000 mc, db; PRO (21) vs Medical/surgical pts with Initial increase in EDTA levels in
PRO EDTA (18) acute or chronic renal PRO EDTA group. ND in vitamin D,
insufficiency requiring parathyroid hormone or ionized
MV calcium levels. No effect on renal
function
Abraham et al.,[200] mc, db; PRO (43) vs Pts with respiratory EDTA levels increased in PRO
2000 PRO EDTA (42) failure/ARDS requiring EDTA group. ND in ionized
MV >48 h calcium, magnesium or parathyroid
hormone levels. No progression to
renal failure
Higgins et al.,[201] mc, ol; PRO EDTA Medical/surgical/trauma Increased urinary excretion of zinc
2000 (106) vs other pts requiring MV and iron in PRO EDTA group. NSD
sedative agent (104) in BUN, Cr or CLCR
Sandiumenge sc, ol; 2% PRO (32) Trauma pts requiring MV ND in sedation adequacy, Higher risk of
et al.,[202] 2000 vs MID (31) >48 h emergence behaviour or hypertriglyceridemia and
haemodynamic variables sedation failure in PRO group
McCollam sc, ol; LOR (10) vs Surgical/trauma pts MID group at target sedation level Particulate precipitation in
et al.,[203] 1999 MID (10) vs requiring MV more often than other groups. LOR intravenous tubing observed in
PRO (10) group more often oversedated. PRO LOR group. Sedation costs
group more often undersedated higher for MID and PRO
Swart et al.,[204] sc, db; LOR (31) vs Medical pts requiring MV Easier to achieve and maintain MID costs substantially higher
1999 MID (33) >3 d target sedation level with LOR. ND
in emergence time
Venn et al.,[205] mc, db, pc; DEX (47) Cardiac and general DEX group required less rescue Higher incidence of persistent
1999 vs PL (51) surgical pts requiring MV MID/morphine bradycardia in DEX group
for >6 h
Kelly et al.,[206] mc, db; PRO (23) vs TBI pts requiring MV PRO group had lower ICP on day 3 ND in mortality or long-term
1999 morphine (19) and less requirement for adjuncts neurological outcome among
such as cerebrospinal fluid survivors
drainage and other sedatives
Carrasco et al.,[207] sc, db; PRO (25) vs Postoperative cardiac ND in duration or adequacy of Hypotension in PRO and MID
1998 MID (25) vs MID and surgical pts requiring MV sedation. PRO groups had shorter groups. Bradycardia in
PRO (25) emergence time and time to combined group. Cost savings
tracheal extubation in combined group

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1895

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Sanchez- sc, ol; MID (34) vs Trauma pts requiring MV ND in achieving target sedation. ND Triglyceride levels higher in
Izquierdo-Riera PRO (33) vs MID >48 h in cerebral haemodynamics. Time PRO-treated pts
et al.,[208] 1998 and PRO (33) to awakening shorter with PRO
Weinbroum sc, ol; PRO (31) vs Medical/surgical pts Sedation quality rated to be better Greater incidence of
et al.,[209] 1997 MID (36) requiring MV >24 h in MID group. Faster degree of hypotension with bolus PRO
recovery with PRO. Higher and more emergence agitation
incidence of amnesia in MID group in PRO group

Hall et al.,[17] 1997 sc, ol; PRO low (10) Postoperative cardiac Light sedation not associated with ND in catecholamine levels
vs high (11) level of surgical pts requiring MV differences in cortisol levels or
sedation for <24 h incidence of myocardial infarction

Searle et al.,[210] sc, db; PRO (21) vs Postoperative CABG ND between groups with respect to ND in requirement for analgesic
1997 MID (20) surgical pts time at target sedation level, or vasoactive agents. NDs in
emergence time or time to tracheal haemodynamic parameters or
extubation measures of pulmonary function

Plunkett et al.,[211] sc, ol; PRO [deep Postoperative CABG Cortisol, epinephrine, dopamine PRO-treated group had lower
1997 sedation] (61) vs surgical pts and norepinephrine levels lower in incidence of tachycardia and
standard care (60) PRO group. Reduced requirements hypertension but higher
for opioids in the PRO group incidence of hypotension

Barrientos-Vega sc, ol; PRO (54) vs Pts requiring MV No differenence in mortality or Higher costs in PRO-treated
et al.,[212] 1997 MID (54) stratified by aetiology ‘therapeutic failure’. PRO-treated group
(cardiorespiratory group had shorter weaning time
failure, trauma,
postoperative or
miscellaneous)

Manley et al.,[213] sc, ol; alfentanil and Pts requiring MV for ND in sedation quality between High number of deaths and
1997 PRO (17) vs >12 h groups. Time to tracheal extubation withdrawals
morphine and and time to be fit for transfer shorter
MID (9) in the alfentanil and PRO group

Treggiari-Venzi sc, ol; PRO (20) vs Conscious non- ND in reported anxiety or

et al.,[214] 1996 MID (20) intubated pts in ICU depression levels between groups
following trauma or post- following overnight sedation
surgical

Chamorro mc, ol; PRO (50) vs Pts (primarily medical) ND in duration of sedation. Better Higher incidence of
et al.,[215] 1996 MID (48) requiring MV for >48 h pt-ventilator synchrony and shorter hypotension in PRO-treated
emergence time in PRO-treated group
group

Wahr et al.,[216] mc, ol; PRO (154) vs CABG pts requiring MV PRO group had less tachycardia ND in mortality
1996 MID (158) for >12 h and hypertension and more
hypotension. ND in incidence or
severity of myocardial ischaemia

Kolenda et al.,[217] sc, ol, TBI pts requiring MV FEN-treated group had lower 4 KET-treated and 5 FEN-
1996 pseudorandomized; requirements for vasoactive agents treated pts withdrawn due to
KET/MID (17) vs and higher cerebral perfusion persistent elevations of ICP
FEN/MID (18) pressure and ICP >25 mm Hg

Cernaianu mc, ol; LOR (50) vs ICU pts requiring MV NSD between groups with respect Larger amounts of MID
et al.,[218] 1996 MID (45) for <8 h to quality of sedation or required to achieve similar
haemodynamics degree of sedation. MID less
cost efficient

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1896 Roberts et al.

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)

Kress et al.,[219] sc, ol; PRO (37) vs Adult medical pts Time to achieve adequate sedation Similar reduction in oxygen
1996 MID (36) requiring MV similar between groups. Emergence consumption between groups
time shorter in PRO group
McArthur et al.,[220] sc, ol; PRO (10) vs Brain-injured pts ND in effects on gastric emptying
1995 morphine/MID (11) requiring MV between groups
Ronan et al.,[221] sc, ol; PRO (30) vs Postoperative surgical Better quality of sedation and more Lower heart rate in PRO group
1995 MID (30) pts requiring MV for rapid emergence in PRO group
>12 h
Levy et al.,[222] sc, blinded; ETO (3) TBI pts requiring MV ND in intra-cranial haemodynamics Study stopped due to
1995 vs pentobarbital (4) development of acute kidney
injury in ETO-treated pts
thought to be secondary to the
ethylene glycol carrier
Higgins et al.,[223] sc, ol; PRO (42) vs Postoperative CABG pts PRO-treated group had lower heart ND in cardiac output. Lower
1994 MID (38) rate and blood pressure on initiation requirement for analgesia and
of therapy vasoactive drugs in PRO-
treated group
Pohlman et al.,[224] sc, ol; LOR (10) vs Medical pts requiring MV ND between groups for time to
1994 MID (10) reach adequate sedation level,
adjustments to maintain sedation
level, duration of emergence or fluid
requirements for infusion
Costa et al.,[225] sc, ol; PRO (164) vs MV ICU pts requiring PRO group had superior quality of In short-term group, time to
1994 MID (153) vs DIA sedation for short sedation. DIA group had poor extubation and recovery were
(35) (<72 h), medium (3 to quality and therefore was not given faster with PRO. In medium
5 d) or long (>5 d) for medium- or long-term sedation duration group, time to
durations extubation and recovery were
shorter with the PRO group. In
the long-term group, time to
extubation and recovery were
faster in PRO group
Carrasco et al.,[226] sc, ol; PRO (46) vs Pts requiring MV PRO-treated pts had greater time PRO had a favourable cost-
1993 MID (42) stratified by short- at target sedation level and shorter benefit analysis in the short-
(<24 h), medium- (24 h to and more predictable emergence term group
7 d), or long- (>7 d) term times
requirement for sedation
Roekaerts sc, ol; PRO (15) vs CABG pts requiring MV Emergence time and time to Hypotension in PRO-treated
et al.,[227] 1993 MID (15) tracheal extubation shorter in PRO group and tachycardia in MID
group group
Boyd et al.,[228] sc, sb; PRO (10) vs Post-abdominal aortic Target sedation level achieved
1993 MID (9) aneurysm surgery pts more frequently. Emergence more
receiving physiotherapy rapid with PRO
while on MV
Chaudhri and sc, ol; PRO (20) vs CABG pts requiring ND in quality of sedation or time to PRO group had higher risk of
Kenny,[229] 1992 MID (20) haemodynamic control tracheal extubation hypotension
and MV
Heinrichs et al.,[230] sc, ol; PRO (30) vs Postoperative pts PRO-treated pts had shorter Hypertriglyceridemia in PRO-
1992 pethidine, requiring MV and recovery times from treated pts
promethazine and sedation for 6 h discontinuation of sedation
dihydroergotamine
(30)

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1897

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Spencer et al.,[231] sc, ol; ISO (30) vs Pts requiring MV for ISO-treated group had more rapid NDs between groups in
1992 MID (30) >24 h emergence and earlier tracheal haemodynamic or biochemical
extubation variables
Beyer and sc, ol; MID (10) vs Postoperative female Shorter recovery period from time Initial hypotension with PRO
Seyde,[232] 1992 PRO (10) pts requiring MV for of sedation discontinuation in PRO bolus accentuated in
>24 h group hypovolaemic pts. Otherwise
ND in haemodynamics.
Epinephrine and
norepinephrine levels lower in
PRO group
Clarke,[233] 1991 sc, ol; PRO (9) vs Postoperative ND between groups for
MID (11) neurosurgical pts haemodynamic parameters or ICP.
receiving MV until first Pts in PRO group had shorter time
postoperative day to tracheal extubation
Boyle et al.,[234] sc, ol; PRO (29) vs Surgical pts requiring ND in time at ‘adequate’ sedation Tolerance to sedative effects
1991 MID (29) MV >12 h level. More requirements for evident in both groups. Drug
infusion adjustments in PRO group. clearance more rapid in PRO
Shorter emergence times for pts group. Hypertryglyceridemia
sedated for >24 h in PRO group more common in PRO group
Beyer and sc, ol; PRO (10) vs Female pts requiring MV PRO group had more hypotension NSD in catecholamine levels
Seyde,[235] 1991 MID (10) following major on induction of sedation. PRO
abdominal surgery group had shorter emergence times
Degauque and sc, ol; PRO (5) vs Pts with COPD requiring Quality of sedation considered More MID-treated pts required
Dupuis,[236] 1991 MID (6) MV superior in PRO group supplemental sedative agents.
ND in ICU LOS
Wolfs et al.,[237] sc, ol; PRO (17) vs Postoperative surgical PRO group had earlier time to ND in quality of sedation
1991 MID (17) pts requiring sedation resumption of spontaneous
for 6 h respiration following
discontinuation of sedative infusion
Adams et al.,[238] sc, ol; FEN (10) vs Pts receiving Lower catecholamine requirements NDs in haemodynamics other
1991 KET (10) catecholamine therapy in KET group than an increase in pulmonary
and MV >48 h artery pressure in the KET
group and an increase in
central venous pressure in the
FEN group
McMurray sc, ol; PRO (50) vs CABG pts requiring MV PRO-treated pts had more rapid Higher requirements for
et al.,[239] 1990 MID (50) emergence and time to tracheal supplemental analgesia in MID-
extubation treated pts
Snellen et al.,[240] sc, ol; PRO (20) vs CABG pts requiring MV PRO-treated pts had more Hypotension in both groups
1990 MID (20) for >12 h rapid emergence and tracheal with initial bolus dose
extubation
Harris et al.,[241] sc, ol; Pts requiring MV ND in degree of satisfactory PRO goup required more
1990 PRO/papaveretum sedation level or haemodynamic or muscle relaxants for ventilator
(15) vs pulmonary variables dysynchrony. No indication of
papaveretum/MID impaired steroidogenesis in
(12) PRO group
Gottardis et al.,[242] sc, ol; PRO (10) vs Trauma/TBI pts ND in levels of triglycerides or
1989 DIA/pentobarbitone requiring MV cholesterol
Kong et al.,[243] sc, ol; ISO (30) vs Critically ill pts requiring ISO group spent more time at target ND in haemodynamics
1989 MID (30) MV for >12 h sedation level with shorter
emergence times

Continued next page

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1898 Roberts et al.

Table IV. Contd

Study Design; Population Main findings Secondary findings
comparators (n)
Boeke et al.,[244] sc, ol; PRO (5) vs Postoperative PRO group had higher level of Hypotension observed in PRO
1989 MID (5) abdominal aortic satisfactory sedation and shorter group and hypertension
surgery pts recovery time following infusion observed in MID group
discontinuation
Aitkenhead mc, ol; PRO (53) vs Pts requiring MV <24 h NSD in duration of time at adequate Heart rate lower in PRO group.
et al.,[245] 1989 MID (47) sedation levels. Time to tracheal ND in mortality. No suppression
extubation shorter in PRO group of adrenocortical function
Ledingham sc, ol; morphine Critically ill/injured pts Morphine infusion and bolus Groups imbalanced at baseline
et al.,[246] 1988 infusion and MID requiring MV for >12 h produced most satisfactory with morphine infusion and MID
bolus (11) vs MID sedation. Pts receiving MID bolus group exhibiting worse
infusion and infusion and morphine bolus had outcomes
morphine bolus (12) least satisfactory sedation
vs morphine infusion (including disorientation)
and morphine bolus
(13)
Adams et al.,[247] sc, ol; FEN (8) vs Surgical pts requiring EEG (Compressed Spectral Array) ND in levels of epinephrine,
1988 KET (8) MV >24 h recordings similar in both groups norepinephrine, antidiuretic
and indicative of sedation. hormone, cortisol, glucose,
Requirements for pancuronium lactate or free glycerol between
lower in KET group groups
Grounds et al.,[248] sc, ol; PRO (30) vs Cardiac surgery pts PRO-treated pts at target sedation
1987 MID (30) requiring MV level longer, and had shorter time to
extubation once infusion
discontinued
Dearden and sc, db; althesin (5) TBI pts requiring MV ND in intracranial haemodynamics One tx failure in each group
McDowall,[249] vs ETO (5)
1985
White et al.,[250] sc, ol; IV FEN (15) vs Pts with diffuse brain FEN produced no change in ICP Cerebral perfusion pressure
1982 IV thiopental (15) vs injury requiring MV and did not blunt the ICP increase not altered in any group
IV lidocaine (15) vs with ETT suctioning. Thiopental
intratracheal and IV lidocaine reduced ICP
lidocaine (15) vs IV initially but failed to blunt the ICP
succinylcholine (15) increase with suctioning
vs IV saline (15) prior
to ETT suctioning
ACTH = adrenocorticotrophic hormone; ARDS = acute respiratory distress syndrome; BIS = Bispectral Index; BUN = blood urea nitrogen;
CABG = coronary artery bypass grafting; COPD = chronic obstructive pulmonary disease; CLCR = creatinine clearance; Cr = creatinine;
db = double blind; dd = double dummy; DEX = dexmedetomidine; DIA = diazepam; EDTA = ethylenediaminetetraacetic acid; EEG = electro-
encephalography; ETO = etomidate; ETT = endotracheal tube; FEN = fentanyl; FOS = fospropofol; ICP = intracranial pressure; ICU = intensive
care unit; IL-6 = interleukin-6; ISO = isoflurane; IV = intravenous; KET = ketamine; LOR = lorazepam; LOS = length of stay; mc = multi-centre;
MID = midazolam; MV = mechanical ventilation; ND = no difference; NSAID = nonsteroidal anti-inflammatory drug; NSD = no statistically
significant difference (p > 0.05); ol = open label; pc = placebo controlled; PL = placebo; PRO = propofol; pt(s) = patient(s); REM = remifentanil;
sb = single blind; sc = single-centre; SEV = sevoflurane; TBI = traumatic brain injury; tx = treatment; VAP = ventilator-associated pneumonia.

(9% of the potency of midazolam).[44] The active diazepines among critically ill patients. Obesity
metabolites of diazepam and midazolam are pre- and hypoalbuminaemia have also been shown to
dominantly renally cleared and may accumulate contribute to a prolonged midazolam effect, likely
during continuous infusions, especially when secondary to an accumulation of the drug in peri-
these are uninterrupted or prolonged.[133] pheral adipose tissue as well as an increase in its
Several factors, including age[262,263] and con- volume of distribution.[44] Co-administration of
comitant diseases and/or drug therapy[51,264,265] af- drugs metabolized by CYP3A4 (e.g. midazolam
fect the intensity and duration of action of benzo- and fluconazole[266]) or CYP2C19 (e.g. diazepam

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1899

and omeprazole[267]) may further increase or pro- may also induce a withdrawal syndrome char-
long the sedative effect. Although disputed,[135] acterized by CNS stimulation and sympathetic
genetic polymorphism may produce variability nervous system upregulation (including gastro-
in midazolam metabolism and response, as in- intestinal hyperactivity) in up to 33% of adult
dividuals homozygous for the CYP3A5*3 and ICU patients.[51]
CYP3A5*6 alleles metabolize the drug slower than As vials of lorazepam contain propylene gly-
those who are homozygous for the CYP3A5*1 col, this drug may produce propylene glycol
allele.[66] Finally, cigarette smoking reduces the toxicity.[269,270] The clinical presentation of pro-
sedative effect of benzodiazepines[263] (likely by pylene glycol toxicity is characterized by meta-
upregulating CYP function) and higher doses of bolic acidosis with an elevated osmolar gap and
benzodiazepines are frequently required when renal dysfunction secondary to reversible acute
patients are regular users of alcohol (ethanol),[44] tubular necrosis.[51,270] Other reported manifesta-
given that these two drugs exhibit significant tions include haemolysis, seizures or coma, CNS
cross-tolerance at the GABAA receptor. depression, agitation, hypotension and cardiac
Benzodiazepine use is associated with several arrhythmias or asystole.[51] As propylene glycol
adverse drug reactions. Lorazepam may result in toxicity is linked with higher doses of lorazepam,
paradoxical agitation, possibly secondary to its and monitoring of propylene glycol plasma con-
ability to produce drug-associated amnesia and/ centrations is impossible or impractical at many
or disorientation or because of ill-defined differ- institutions, an osmol gap (in combination with
ences in its pharmacodynamics as compared with the results of arterial blood gases) should be fol-
midazolam or diazepam.[51] Moreover, high con- lowed when the total lorazepam dose exceeds
centrations of lorazepam may precipitate in intra- 1 mg/kg/day.[51] If toxicity is suspected, lorazepam
venous administration tubing. To prevent this, should be discontinued, supportive therapy in-
the drug must be diluted to a concentration of itiated and an alternate sedative agent that does
<1 mg/10 mL prior to administering it as a con- not contain propylene glycol used as a replace-
tinuous infusion.[12,268] Acute discontinuation of ment. Haemodialysis can be utilized in severe
benzodiazepines after prolonged administration cases of toxicity.[51]

Table V. Systematic reviews and/or meta-analyses of sedative agents in the intensive care unit
Study Study design and Main findings
comparators
(no. of included studies)
Ostermann RCTs of sedative agents (32) Propofol as effective as midazolam with shorter time to tracheal extubation. Propofol
et al.,[14] 2000 produced more hypotension and was more expensive than midazolam. Isoflurane
had some advantages over midazolam while ketamine was more favourable than
fentanyl for pts with traumatic brain injury
Walder et al.,[251] RCTs of propofol vs Duration of adequate sedation longer with propofol. In postoperative pts receiving
2001 midazolam (27) mechanical ventilation for <36 h, weaning was faster with propofol. Incidence of
hypotension and hypertriglyceridaemia higher in propofol-treated pts
Magarey,[252] 2001 RCTs of propofol vs Quality of sedation inconclusive. Time to tracheal extubation and recovery times
midazolam (17) were shorter with propofol. Propofol more likely to produce hypotension and
bradycardia
Ho and Ng,[253] RCTs of propofol vs alternative No difference in mortality between propofol and alternative sedative agents. Use of
2008 sedative agents (16) propofol for medium and long duration of sedation resulted in significantly shorter
ICU LOS. However, when compared with midazolam alone, there was NSD in LOS
Roberts et al.,[2] RCTs of sedative and No evidence exists that any single sedative agent is associated with improved
2011 analgesic agents for critically ill neurological outcomes or death among ICU pts with severe traumatic brain injury.
adults with severe traumatic Bolus doses of opioids may transiently increase intracranial pressure and reduce
brain injury (13) cerebral perfusion pressure
ICU = intensive care unit; LOS = length of stay; NSD = no statistically significant difference; pts = patients; RCT = randomized controlled trial.

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1900 Roberts et al.

9.2 Anaesthetic Agents bean oil, 2.25% glycerol and 1.2% purified egg
phosphatide), which may lead to injection site
9.2.1 Propofol pain in up to two-thirds of patients after admin-
Propofol (2,6-diisopropylphenol) is a sterically istration into smaller peripheral veins.[272] In ad-
hindered, highly lipophilic phenol classified as an dition, its formulation supports bacterial growth,
ultra-short-acting intravenous anaesthetic agent. and therefore lapses in sterile technique increase
This drug is also available as a phosphorylated the risk of bacterial infection.[284] The manu-
pro-drug named fospropofol. Propofol acts on the facturer therefore suggests that storage bottles
GABAA receptor[271,272] and, like the benzodiaze- and intravenous line tubing be discarded every
pines, exhibits dose-dependent sedation ranging 12 hours to maintain line integrity and minimize
from mild depression of responsiveness to ob- risk of bacterial contamination. Similar to benzo-
tundation. It has powerful anxiolytic and am- diazepines, propofol causes respiratory depression,
nestic properties.[273] Importantly, propofol has which may lead to apnoea.[285] The cardiovascular
no analgesic effect. In fact, in one comparative effects of propofol include decreases in blood pres-
study, patients receiving propofol required even sure, especially among hypovolemic patients, which
higher doses of morphine than those sedated with is secondary to reduced preload, increased venous
midazolam, highlighting that appropriate analge- capacitance, decreased venous return and mild de-
sia is required regardless of the type of sedation pression of myocardial contractility.[278,286,287]
employed.[219] Propofol can also be used as an an- Propofol may also cause hyperlipidaemia.
ticonvulsant, and may be effective for treatment of Although the incidence of hyperlipidaemia is
refractory status epilepticus.[274,275] largely unknown, one retrospective cohort study
Propofol is highly lipophilic and rapidly crosses reported its development in 18% of 512 propofol-
the blood-brain barrier following administration, treated critically ill patients.[288] As hypertriglycer-
resulting in a rapid onset (1–5 minutes) of sedation, idaemia may lead to pancreatitis, patients receiving
amnesia and hypnosis.[182,276] Its duration of ac- propofol should have their serum triglyceride le-
tion is dose dependent[277] and relatively short vels monitored closely if propofol is administered
(2–8 minutes for a bolus injection), highlighting for 48 hours or longer.[288] If hyperlipidaemia de-
the rapidity of its redistribution out of brain and velops, the drug should be discontinued. Hy-
into peripheral tissues.[278] This rapid movement perlipidaemia is typically associated with high
between compartments may explain the observa- infusion rates (3–6 mg/kg/h), concurrent admin-
tion in RCTs and systematic reviews that the drug istration of parenteral lipids and the presence of
often affords more accelerated recovery from se- baseline hyperlipidaemia. Parenteral lipid nutri-
dation than midazolam. tion should be adjusted to account for the calories
Propofol undergoes phase II metabolism in administered, as commercial 1% propofol affords
the liver into inactive glucuronide and sulphate 1.1 kcal/mL of energy.[289]
metabolites that are renally cleared.[279] How- Propofol infusion syndrome (PRIS) is a rare
ever, unlike midazolam and diazepam, the me- complication of propofol administration.[290,291]
tabolism of propofol is not entirely dependent on Although its pathophysiology is incompletely
hepatic metabolism, which is evidenced by its understood, this potentially fatal syndrome may
continued clearance during the anhepatic phase result from mitochondrial respiratory chain dys-
of liver transplantation.[280-282] Despite this, elderly function.[290,292,293] Risk factors for PRIS include
patients require a dose reduction due to their de- propofol infusion rates >5 mg/kg/h for >48 hours,
creased volume of distribution, which results in a utilization of higher concentrations (2% vs 1%)
sustained and higher systemic concentration of the of propofol, young age, critical illness, high fat
drug and less peripheral redistribution.[278,283] and low carbohydrate intake, inborn errors of
Propofol use has been associated with several mitochondrial fatty acid oxidation and con-
adverse drug reactions. The drug is formulated as comitant catecholamine infusion or steroid ther-
an oil-in-water emulsion (containing 10% soya apy.[290-295] While usually observed in patients

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1901

receiving high doses of propofol for prolonged The AnaConDa device is a reflection filter (a
periods of time, it may also occur after a short device connected between the ventilator and pa-
infusion and/or with low doses.[296-298] Clinically, tient that permits re-inhalation of anaesthetics)
patients with PRIS develop metabolic acidosis that allows for infusion of liquid volatile anaes-
and refractory bradycardia leading to asystole, thetics into the ventilator circuit via a syringe
hyperkalaemia, rhabdomyolysis, hyperlipidaemia pump.[306] This device uses charcoal as an ab-
and/or an enlarged fatty liver. Monitoring for sorbing agent and has an imbedded heat and
syndrome development consists of following sys- moisture exchanger.[306] Although the AnaConDa
temic lactate levels,[299] mixed venous oxygen has been introduced for use in the ICU set-
saturations and serum triglycerides. Treatment ting[300,302] (table IV), its widespread uptake is
consists of immediate discontinuation of pro- likely limited secondary to various device- and
pofol, correction of haemodynamic and meta- non-device-related concerns.[301,306] First, chang-
bolic abnormalities and possibly use of cardiac ing the respiratory rate or tidal volume altered
pacing.[278] the expired volatile anaesthetic fraction in one
bench study, and the standard luer lock on the
9.2.2 Volatile Inhalational Agents liquid volatile-filled syringe allowed it to be in-
Concerns regarding the adverse effects of in- advertently connected to an intravenous infusion
travenous sedative agents, coupled with the devel- line.[306] Second, there may be issues surrounding
opment and release of the Anesthesia Conserving environmental pollution and adequate scavenging
Device (AnaConDa, Hudson RCI, Uppsland of waste gas.[300,301,307] Non-anaesthesia-trained
Väsby, Sweden), have fostered renewed interest intensivists may also be unfamiliar with use of vol-
into use of volatile inhalational anaesthetics (e.g. atile anaesthetics, including the concept of MAC,
desflurane, isoflurane and sevoflurane) for ICU and their introduction into ICU practice could
sedation.[300-302] At low alveolar concentrations, have significant safety, cost and environmental
volatile anaesthetics produce amnesia, euphoria, implications.[301,307] Fourth, most studies have
analgesia and hypnosis.[50] At higher concentra- examined their use for relatively short durations of
tions, they lead to deep sedation, muscle relaxation administration (hours), and the safety and toxicity
and diminished motor and autonomic responses to of administration for a more prolonged period of
painful stimuli.[50] time among patients with varying degrees of organ
Volatile anaesthetic agents have several char- dysfunction remains uncertain.[300,301,304,308-309]
acteristics that may be advantageous when used as Finally, malignant hyperthermia is a specific and
sedative agents. They have a quick and consistent potentially fatal adverse drug event precipitated
onset and offset of action, and may be easily and by volatile anaesthetic agent administration and
rapidly titrated upward and downward until the education regarding its clinical features and a
desired level of sedation is achieved.[47,168,303-305] management plan devoted to treating the condi-
Moreover, although metabolism of volatile agents tion should be developed before sedation with
occurs to some extent,[303] modern volatile anaes- volatile anaesthetic agents is introduced.[310]
thetics such as desflurane, sevoflurane and isoflu-
rane predominantly undergo pulmonary clearance 9.2.3 Etomidate
and thus are less affected by liver or kidney dis- Etomidate is a GABAA receptor agonist used as
ease/failure.[47] In addition, the sedative effect of an induction agent for rapid sequence intubation in
volatile anaesthetic gases may be followed via use critically ill patients.[48] The major clinical advantage
of the median alveolar concentration (MAC), of etomidate is its favourable cardiorespiratory
with one MAC being defined as ‘‘the end-tidal effects. At intubation doses, it produces little to
concentration (in standardized units) of inhaled no change in heart rate or mean arterial pres-
anaesthetic that ablates movement (e.g. withdrawal) sure.[48,311-313] Moreover, the drug decreases cere-
in response to surgical incision in 50 percent of a test bral blood flow, cerebral metabolic rate and ICP
population.’’[50] among patients with severe traumatic brain in-

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1902 Roberts et al.

jury.[255] Thus, it may frequently be used to fac- cardiac output, it may be useful for the sedation
ilitate intubation in multisystem trauma patients of patients following trauma[191,325,326] or with
with suspected neurological injury and those with septic shock.[327,328] Further, because synaptic
haemorrhagic or septic shock.[314,315] NMDA receptors increase and GABAA recep-
Unfortunately, etomidate causes adrenal sup- tors decrease with time during prolonged sei-
pression as the drug is a potent inhibitor of the zures, ketamine may prove useful for refractory
adrenal cortical enzyme 11b-hydroxylase, which status epilepticus in future studies.[329]
converts 11b-deoxycortisol into cortisol.[316] A Unfortunately, the use of ketamine in critical
recent systematic review and meta-analysis of care medicine is limited by its adverse drug reactions.
14 randomized and non-randomized studies re- These include hallucinations during dissociative
ported that a bolus dose of etomidate was asso- anaesthesia, emergence delirium, dissociative or
ciated with not only an increased risk of adrenal out-of-body experiences, unpleasant recall effects
suppression (relative risk [RR] 1.64; 95% CI 1.52, or flashbacks, hypersalivation, lacrimation, tachy-
1.77), but also a heightened incidence of mor- cardia and increased myocardial oxygen de-
tality (RR 1.19; 95% CI 1.10, 1.30).[52] Thus, mand.[44,47,49,330] Although ketamine has also
some have called for the use of etomidate for been associated with an increase in ICP in several
endotracheal intubation in critically ill patients to earlier clinical reports,[331-333] a recent systematic
be limited or even abandoned.[52,317] review reported that ketamine had largely neutral
effects on ICP and CPP among critically injured
9.2.4 Ketamine adults with severe traumatic brain injury.[2] More-
Ketamine is an arylcyclohexylamine neuro- over, no randomized evidence was found to sug-
leptic anaesthetic agent.[49,318,319] It blocks the gest that ketamine was associated with a higher
glutamatergic N-methyl-D-aspartate (NMDA) risk of unfavourable neurological outcomes or an
receptor,[320] inhibits cellular calcium influx and elevated mortality among those with severe head
produces dose-dependent amnesia, analgesia and injury.[2]
sedation.[49] Ketamine induces a state of ‘dis-
sociative anaesthesia’, during which patients are 9.3 a2-Adrenergic Receptor Agonists
conscious and will remain breathing spontaneously 9.3.1 Clonidine
with maintained eye-opening and reflexes, includ- Clonidine is a centrally acting a2-adrenergic
ing laryngeal reflexes.[49] receptor agonist.[334] Originally introduced as an
Ketamine is highly lipid-soluble, and reaches antihypertensive agent,[334,335] it has sedative-
maximal effect within 1 minute of administration and analgesic-sparing properties.[336] It may have
of an intravenous bolus dose.[47,319] Thereafter, its utility as an adjunct for sedation of ICU patients
duration of action is only 10–15 minutes.[47,319] The with a history of opioid[337] or alcohol[338,339] abuse
drug is extensively metabolized in the liver by and/or to ameliorate withdrawal syndromes fol-
CYP2B6, CYP2C8/9 and CYP3A4 to norketa- lowing prolonged sedative and narcotic adminis-
mine (which is one-fifth to one-third as potent tration in the ICU.[340] Common adverse effects of
as ketamine) and hydroxyl-norketamine, both of clonidine include hypotension, bradycardia and
which are subsequently renally cleared.[318] xerostomia.[334] Sudden discontinuation of the agent
Ketamine activates the sympathetic nervous after prolonged use may also be associated with re-
system, maintains mean arterial pressure, increases bound hypertension and tachycardia in susceptible
heart rate and produces bronchodilation.[44,49] patients.[341,342] A significant limitation of clonidine
Animal models and in vitro studies have also is the lack of an intravenous formulation in North
demonstrated that the drug has anti-inflamma- America.
tory effects as it inhibits production of several
pro-inflammatory cytokines, including tumour 9.3.2 Dexmedetomidine
necrosis factor-a.[321-324] As ketamine increases or Dexmedetomidine is a centrally acting, highly
at least maintains mean arterial pressure and selective, a2-adrenergic receptor agonist that pos-

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1903

sesses both sedative and analgesic properties.[164] It volemia (e.g. secondary to truncal trauma-related
is currently approved in Canada and the US for haemorrhage or severe sepsis).
short-term sedation (<24 hours) of patients re- As dexmedetomidine may induce hypotension
ceiving mechanical ventilation. Despite this, RCTs and bradycardia, a bolus or initial loading dose is
that administered dexmedetomidine for longer seldom used in practice.[346] Instead, the drug is
durations have been conducted and the drug has slowly titrated up to 0.2–0.7 mg/kg/hour. There-
not yet demonstrated any significant safety con- after, the dosage is increased every 30 minutes
cerns (table IV).[17,23,26,159,163,165-167,169-250] until the desired clinical effects are achieved.
Dexmedetomidine has sedative, analgesic and There may be a (as yet undefined) plateau effect
anxiolytic effects. However, unlike other sedative to drug dosing, beyond which little additional
agents, dexmedetomidine is not associated with sedative effects are observed.[347,348] Though there
respiratory depression, a feature that may ulti- are no specific guidelines for modifying the dose
mately prove to be highly useful for critically ill for elderly patients or those with hepatic impair-
patients.[343] There is also evidence that dexmede- ment, it is advisable to start at the low end of the
tomidine decreases the need for alternative seda- dosage range and titrate up slowly until the de-
tives,[205] and may increase the number of days sired degree of sedation is achieved.
without delirium[179] and decrease time to extuba-
tion when compared with the benzodiazepines.[159] 10. Pharmacoeconomic Analyses
The pharmacokinetics of dexmedetomidine are
well described. The drug has a distribution half-life The provision of critical care to mechanically
of approximately 6 minutes following intravenous ventilated ICU patients is very costly.[349] While
administration, and an estimated terminal elim- the individual costs of the agents required to
ination half-life of 2 hours. Dexmedetomidine is provide sedation are not generally excessive, the
completely metabolized through phase I CYP2A6- aggregate costs for provision of sedation and
mediated hydroxylation and phase II direct analgesia to permit tolerance of mechanical venti-
N-glucuronidation to inactive metabolites.[344] lation are significant. The requirement for sedative
Although age, sex and renal impairment do not agent administration is therefore one of the drivers
affect the pharmacokinetics of the drug, hepatic for the increased costs of care beyond that required
impairment reduces its clearance significantly.[344] for non-ventilated patients.[349] Attempts to mini-
Among patients with mild, moderate and severe mize time spent on mechanical ventilation may
hepatic impairment, mean clearance of the drug is therefore be seen as a cost-saving manoeuver, ir-
approximately 74%, 64% and 53% of that ob- respective of any clinical benefits derived from this
served in healthy volunteers, respectively.[344] practice. However, any analysis of the costs of the
The most frequently observed adverse effects provision of sedation for this purpose must con-
of dexmedetomidine in RCTs include xerostomia, sider several factors when trying to interpret stud-
bradycardia and hypotension (tables III and ies exploring the cost-effectiveness analysis of one
IV[17,23,26,159,163,165-167,169-250]). The associated sedative regime over another.
cardiovascular effects of dexmedetomidine are First is the acquisition cost of the agents in-
complex and likely result in a biphasic blood volved. This is often seen as a significant barrier
pressure response. Immediately following a bolus to the introduction of new agents (e.g. dexmede-
injection, bradycardia and hypertension occur, tomidine) into ICU practice. However, several
perhaps due to vasoconstriction from activation additional factors should be considered. For ex-
of peripheral a2b-adrenergic receptors while hypo- ample, if the agent produces fewer side effects
tension is observed during continuous infusions (e.g. delirium), it may be that the cost increase
due to central sympatholysis and activation of becomes neutralized by the costs of management
central a2a-adrenergic receptors.[345] Importantly, of the side effects of cheaper regimens (e.g. prolonged
this hypotensive effect may be more pronounced ICU stay or duration of mechanical ventilation,
among critically ill adults with pre-existing hypo- management of ventilator-acquired pneumonia due

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1904 Roberts et al.

to over-sedation, missed complications such as me- targeted sedation and earlier time to tracheal
senteric ischaemia and requirements for more di- extubation with propofol than midazolam, de-
agnostic tests [e.g. computed tomography], among spite no apparent difference in cost.[351] We at-
others).[350] tributed this to the realities of the healthcare
The second consideration is the healthcare system in Canada, wherein there may not be a
system in which the study was conducted. In a bed into which a discharged ICU patient can be
multicentre study performed in Canada, our admitted. Thus, the prolonged ICU stay effec-
group was able to demonstrate better levels of tively eliminates any differential cost benefits

Table VI. Economic evaluations of alternative sedation regimes in the intensive care unit
Economic evaluation Regimens or agents compared Main findings
Dasta et al.,[352] 2010 Midazolam vs dexmedetomidine Dexmedetomidine produced significant cost savings
Al et al.,[353] 2010 Conventional sedation vs Remifentanil and propofol were more cost effective than conventional
remifentanil and propofol sedation
Cox et al.,[354] 2008 Propofol vs lorazepam or Propofol was less costly than lorazepam but not vs midazolam
midazolam
Panharipande et al.,[179] Lorazepam vs dexmedetomidine No cost difference between groups
2007
Dasta et al.,[355] 2006 Midazolam and propofol vs The three drug combination produced significant cost savings
midazolam, propofol and
dexmedetomidine
Richman et al.,[181] 2006 Midazolam vs midazolam and No cost difference between groups
fentanyl
Muellejans et al.,[356] 2006 Midazolam/fentanyl vs No cost difference between groups
propofol/remifentanil
MacLaren and Sullivan,[357] Midazolam vs lorazepam vs Propofol most cost effective for short-term sedation (<24 h), midazolam
2005 propofol for intermediate sedation (24–72 h) and lorazepam for long-term
sedation (>72 h)
Anis et al.,[351] 2002 Propofol vs midazolam No cost difference between groups
Barrientos-Vega et al.,[358] Midazolam vs 1% propofol vs 2% Compared with historical controls (propofol or midazolam), propofol
2001 propofol groups were more cost effective than midazolam
McCollam et al.,[203] 1999 Midazolam vs lorazepam vs Lorazepam was the most cost effective, followed by midazolam and
propofol then propofol
Swart et al.,[204] 1999 Midazolam vs lorazepam Lorazepam more cost effective than midazolam
Carrasco et al.,[207] 1998 Midazolam vs propofol vs The synergistic combination produced more cost savings than other
midazolam/propofol groups
Searle et al.,[210] 1997 Midazolam vs propofol Infusion costs of propofol were higher than for midazolam
Devlin et al.,[4] 1997 Protocol-driven vs non-protocol- Protocol-driven regimens were less costly
driven sedation regimens
Barrientos-Vega et al.,[212] Midazolam vs propofol Per-patient cost of midazolam was higher than for propofol
1997
Manley et al.,[213] 1997 Alfentanil and propofol vs Alfentanil and propofol was more cost effective than morphine and
morphine and midazolam midazolam
Cernaianu et al.,[218] 1996 Midazolam vs lorazepam Lorazepam was more cost effective than midazolam
Costa et al.,[225] 1994 Midazolam vs propofol vs diazepam The diazepam group was the most costly. For short-term use (<72 h),
propofol was most cost effective. For medium-term use (3–5 d), there
was no difference between propofol and midazolam. For long-term use
(>5 d), there was no difference in cost vs midazolam
Carrasco et al.,[226] 1993 Midazolam vs propofol For sedation <24 h, propofol was more cost effective. However, there
was no difference in costs between midazolam and propofol for
medium- (<24 h to 7 d) or long-term (>7 d) use

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1905

provided by the reduced time to tracheal ex- introduction of sedation and pain scales. Am J Crit Care
tubation. Table VI affords an overview of studies 2008 Jul; 17 (4): 349-56
where costs were considered in the overall anal- 6. Masica AL, Girard TD, Wilkinson GR, et al. Clinical se-
dation scores as indicators of sedative and analgesic drug
ysis.[4,179,181,203,204,207,210,212,213,218,225,226,351-358] exposure in intensive care unit patients. Am J Geriatr
Pharmacother 2007 Sep; 5 (3): 218-31
11. Conclusions 7. Martin J, Franck M, Sigel S, et al. Changes in sedation
management in German intensive care units between 2002
and 2006: a national follow-up survey. Crit Care 2007; 11
Many patients admitted to the ICU will require (6): R124
sedation at some point during their stay. How- 8. Mehta S, McCullagh I, Burry L. Current sedation practices:
ever, emerging evidence suggests that the use of lessons learned from international surveys. Anesthesiol
Clin 2011 Dec; 29 (4): 607-24
sedation should be limited and patients transi-
9. Jackson DL, Proudfoot CW, Cann KF, et al. The incidence
tioned to an awake state as soon as safely possible. of sub-optimal sedation in the ICU: a systematic review.
Knowledge of factors influencing drug effect in Crit Care 2009; 13 (6): R204
critically ill or injured patients, adverse outcomes 10. Marshall J, Finn CA, Theodore AC. Impact of a clinical
associated with prolonged or uninterrupted se- pharmacist-enforced intensive care unit sedation protocol
on duration of mechanical ventilation and hospital stay.
dation and individual drug pharmacodynamics Crit Care Med 2008 Feb; 36 (2): 427-33
and pharmacokinetics will enable the practitioner 11. Woien H, Stubhaug A, Bjork IT. Analgesia and sedation of
to tailor the sedation regime to meet pre-defined mechanically ventilated patients: a national survey of clin-
ical practice. Acta Anaesthesiol Scand 2012 Jan; 56 (1): 23-9
goals and achieve an adequate level of sedation.
12. Jacobi J, Fraser GL, Coursin DB, et al. Clinical practice
guidelines for the sustained use of sedatives and analgesics
in the critically ill adult. Crit Care Med 2002 Jan; 30 (1):
Acknowledgements 119-41
13. Martin J, Heymann A, Basell K, et al. Evidence and con-
Dr Roberts is supported by an Alberta Innovates – Health sensus-based German guidelines for the management of
Solutions (AIHS) Clinician Fellowship Award and research analgesia, sedation and delirium in intensive care–short
funding from the Clinician Investigator and Surgeon Scientist version. Ger Med Sci 2010 Feb 2; 8: Doc02
Programs at the University of Calgary.
14. Ostermann ME, Keenan SP, Seiferling RA, et al. Sedation
Dr Hall is supported by grants from the Nova Scotia in the intensive care unit: a systematic review. JAMA 2000
Heart and Stroke Foundation and the Canadian Anesthe- Mar 15; 283 (11): 1451-9
siologists’ Society RA Gordon Patient Safety Award.
15. Goodwin H, Lewin JJ, Mirski MA. ‘Cooperative sedation’:
These funding organizations had no role in the design of
optimizing comfort while maximizing systemic and
the review; collection, management, analysis, or interpreta-
neurological function. Crit Care 2012; 16 (2): 217
tion of the data; the writing of the report; or the decision to
submit the paper for publication. The authors have no con- 16. Michalopoulos A, Nikolaides A, Antzaka C, et al. Change
flicts of interest to declare. in anaesthesia practice and postoperative sedation shortens
ICU and hospital length of stay following coronary artery
bypass surgery. Respir Med 1998 Aug; 92 (8): 1066-70
References 17. Hall RI, MacLaren C, Stafford Smith M, et al. Light versus
1. Hansen-Flaschen JH, Brazinsky S, Basile C, et al. Use of heavy sedation after cardiac surgery: myocardial ischemia
sedating drugs and neuromuscular blocking agents in patients and the stress response. Maritime Heart Centre and Dal-
requiring mechanical ventilation for respiratory failure: a housie University. Anesth Analg 1997 Nov; 85 (5): 971-8
national survey. JAMA 1991 Nov 27; 266 (20): 2870-5 18. Kress JP, Pohlman AS, O’Connor MF, et al. Daily inter-
2. Roberts DJ, Hall RI, Kramer AH, et al. Sedation for cri- ruption of sedative infusions in critically ill patients un-
tically ill adults with severe traumatic brain injury: a dergoing mechanical ventilation. N Engl J Med 2000 May
systematic review of randomized controlled trials. Crit 18; 342 (20): 1471-7
Care Med 2011 Dec; 39 (12): 2743-51 19. Kress JP, Gehlbach B, Lacy M, et al. The long-term psy-
3. Brook AD, Ahrens TS, Schaiff R, et al. Effect of a nursing- chological effects of daily sedative interruption on criti-
implemented sedation protocol on the duration of me- cally ill patients. Am J Respir Crit Care Med 2003 Dec 15;
chanical ventilation. Crit Care Med 1999 Dec; 27 (12): 168 (12): 1457-61
2609-15 20. Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safety
4. Devlin JW, Holbrook AM, Fuller HD. The effect of ICU of a paired sedation and ventilator weaning protocol for
sedation guidelines and pharmacist interventions on clin- mechanically ventilated patients in intensive care (Awa-
ical outcomes and drug cost. Ann Pharmacother 1997 kening and Breathing Controlled trial): a randomised
Jun; 31 (6): 689-95 controlled trial. Lancet 2008 Jan 12; 371 (9607): 126-34
5. Williams TA, Martin S, Leslie G, et al. Duration of me- 21. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early
chanical ventilation in an adult intensive care unit after physical and occupational therapy in mechanically ventilated,

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1906 Roberts et al.

critically ill patients: a randomised controlled trial. Lancet patients: an observational study. Acta Anaesthesiol Scand
2009 May 30; 373 (9678): 1874-82 2012 May; 56 (5): 645-54
22. de Wit M, Gennings C, Jenvey WI, et al. Randomized trial 40. Mei W, Seeling M, Franck M, et al. Independent risk fac-
comparing daily interruption of sedation and nursing- tors for postoperative pain in need of intervention early
implemented sedation algorithm in medical intensive care after awakening from general anaesthesia. Eur J Pain
unit patients. Crit Care 2008; 12 (3): R70 2010 Feb; 14 (2): 149-7
23. Carson SS, Kress JP, Rodgers JE, et al. A randomized trial 41. Blaise GA, Nugent M, McMichan JC, et al. Side effects
of intermittent lorazepam versus propofol with daily in- of nalbuphine while reversing opioid-induced respiratory
terruption in mechanically ventilated patients. Crit Care depression: report of four cases. Can J Anaesth 1990 Oct;
Med 2006 May; 34 (5): 1326-32 37 (7): 794-7
24. Mehta S, Burry L, Martinez-Motta JC, et al. A randomized 42. Puntillo KA. Pain experiences of intensive care unit
trial of daily awakening in critically ill patients managed patients. Heart Lung 1990 Sep; 19 (5 Pt 1): 526-33
with a sedation protocol: a pilot trial. Crit Care Med 2008 43. Turner JS, Briggs SJ, Springhorn HE, et al. Patients’ re-
Jul; 36 (7): 2092-9 collection of intensive care unit experience. Crit Care Med
25. Kress JP, Vinayak AG, Levitt J, et al. Daily sedative inter- 1990 Sep; 18 (9): 966-8
ruption in mechanically ventilated patients at risk for cor- 44. Wagner BK, O’Hara DA. Pharmacokinetics and pharma-
onary artery disease. Crit Care Med 2007 Feb; 35 (2): 365-71 codynamics of sedatives and analgesics in the treatment of
26. Strøm T, Martinussen T, Toft P. A protocol of no sedation agitated critically ill patients. Clin Pharmacokinet 1997
for critically ill patients receiving mechanical ventilation: Dec; 33 (6): 426-53
a randomised trial. Lancet 2010 Feb 6; 375 (9713): 475-80 45. Devlin JW, Roberts RJ. Pharmacology of commonly used
27. Strøm T, Stylsvig M, Toft P. Long-term psychological ef- analgesics and sedatives in the ICU: benzodiazepines,
fects of a no-sedation protocol in critically ill patients. Crit propofol, and opioids. Anesthesiol Clin 2011 Dec; 29 (4):
Care 2011; 15 (6): R293 567-85
28. Sessler CN, Pedram S. Protocolized and target-based se- 46. Power BM, Forbes AM, van Heerden PV, et al. Pharma-
dation and analgesia in the ICU. Anesthesiol Clin 2011 cokinetics of drugs used in critically ill adults. Clin Phar-
Dec; 29 (4): 625-50 macokinet 1998 Jan; 34 (1): 25-56
29. Pipeling MR, Fan E. Therapies for refractory hypoxemia in 47. Panzer O, Moitra V, Sladen RN. Pharmacology of seda-
acute respiratory distress syndrome. JAMA 2010 Dec 8; tive-analgesic agents: dexmedetomidine, remifentanil,
304 (22): 2521-7 ketamine, volatile anesthetics, and the role of peripheral
30. Brodie D, Bacchetta M. Extracorporeal membrane oxyge- Mu antagonists. Anesthesiol Clin 2011 Dec; 29 (4): 587-
nation for ARDS in adults. N Engl J Med 2011 Nov 17; 605, vii
365 (20): 1905-14 48. Forman SA. Clinical and molecular pharmacology of eto-
31. Sanborn TA, Feldman T. Management strategies for cardio- midate. Anesthesiology 2011 Mar; 114 (3): 695-707
genic shock. Curr Opin Cardiol 2004 Nov; 19 (6): 608-12 49. White PF, Way WL, Trevor AJ. Ketamine: its pharma-
32. Brooks DE, Levine M, O’Connor AD, et al. Toxicology in cology and therapeutic uses. Anesthesiology 1982 Feb; 56
the ICU: part 2. Specific toxins. Chest 2011 Oct; 140 (4): (2): 119-36
1072-85 50. Campagna JA, Miller KW, Forman SA. Mechanisms of
33. Levine M, Ruha AM, Graeme K, et al. Toxicology in the actions of inhaled anesthetics. N Engl J Med 2003 May 22;
ICU: part 3. Natural toxins. Chest 2011 Nov; 140 (5): 348 (21): 2110-24
1357-70 51. Devlin JW, Mallow-Corbett S, Riker RR. Adverse drug
34. Kahn JM, Andersson L, Karir V, et al. Low tidal volume events associated with the use of analgesics, sedatives, and
ventilation does not increase sedation use in patients with antipsychotics in the intensive care unit. Crit Care Med
acute lung injury. Crit Care Med 2005 Apr; 33 (4): 766-71 2010 Jun; 38 (6 Suppl.): S231-43
35. Cheng IW, Eisner MD, Thompson BT, et al. Acute effects 52. Albert SG, Ariyan S, Rather A. The effect of etomidate on
of tidal volume strategy on hemodynamics, fluid balance, adrenal function in critical illness: a systematic review.
and sedation in acute lung injury. Crit Care Med 2005 Jan; Intensive Care Med 2011 Jun; 37 (6): 901-10
33 (1): 63-70 53. Carter CR, Kozuska JL, Dunn SM. Insights into the
36. Jones JG. Perception and memory during general anaes- structure and pharmacology of GABA(A) receptors.
thesia. Br J Anaesth 1994 Jul; 73 (1): 31-7 Future Med Chem 2010 May; 2 (5): 859-75
37. Jones C, Griffiths RD, Humphris G, et al. Memory, delu- 54. Hara M, Kai Y, Ikemoto Y. Enhancement by propofol of
sions, and the development of acute posttraumatic stress the gamma-aminobutyric acid: a response in dissociated
disorder-related symptoms after intensive care. Crit Care hippocampal pyramidal neurons of the rat. Anesthesiol-
Med 2001 Mar; 29 (3): 573-80 ogy 1994 Oct; 81 (4): 988-94
38. Samuelson KA, Lundberg D, Fridlund B. Light vs. heavy 55. Amrein R, Hetzel W. Pharmacology of drugs frequently
sedation during mechanical ventilation after oesopha- used in ICUs: midazolam and flumazenil. Intensive Care
gectomy: a pilot experimental study focusing on memory. Med 1991; 17 Suppl. 1: S1-10
Acta Anaesthesiol Scand 2008 Sep; 52 (8): 1116-23 56. Salgado H, Garcia-Oscos F, Martinolich L, et al. Pre-
39. Jeitziner MM, Schwendimann R, Hamers JP, et al. As- and postsynaptic effects of norepinephrine on gamma-
sessment of pain in sedated and mechanically ventilated aminobutyric acid-mediated synaptic transmission in

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1907

layer 2/3 of the rat auditory cortex. Synapse 2012 Jan; 66 interaction between midazolam and clarithromycin. Clin
(1): 20-8 Pharmacol Ther 1998 Aug; 64 (2): 133-43
57. Seyrek M, Halici Z, Yildiz O, et al. Interaction between 74. Hiller A, Olkkola KT, Isohanni P, et al. Unconsciousness
dexmedetomidine and alpha-adrenergic receptors: em- associated with midazolam and erythromycin. Br J Anaesth
phasis on vascular actions. J Cardiothorac Vasc Anesth 1990 Dec; 65 (6): 826-8
2011 Oct; 25 (5): 856-62 75. Backman JT, Olkkola KT, Neuvonen PJ. Rifampin dras-
58. De Paepe P, Belpaire FM, Buylaert WA. Pharmacokinetic tically reduces plasma concentrations and effects of oral
and pharmacodynamic considerations when treating midazolam. Clin Pharmacol Ther 1996 Jan; 59 (1): 7-13
patients with sepsis and septic shock. Clin Pharmacokinet 76. Tortorici MA, Kochanek PM, Poloyac SM. Effects of
2002; 41 (14): 1135-51 hypothermia on drug disposition, metabolism, and re-
59. Malacrida R, Fritz ME, Suter PM, et al. Pharmacokinetics sponse: a focus of hypothermia-mediated alterations on
of midazolam administered by continuous intravenous the cytochrome P450 enzyme system. Crit Care Med 2007
infusion to intensive care patients. Crit Care Med 1992 Sep; 35 (9): 2196-204
Aug; 20 (8): 1123-6 77. Fukuoka N, Aibiki M, Tsukamoto T, et al. Biphasic concen-
60. Shafer A, Doze VA, White PF. Pharmacokinetic variability tration change during continuous midazolam administration
of midazolam infusions in critically ill patients. Crit Care in brain-injured patients undergoing therapeutic moderate
Med 1990 Sep; 18 (9): 1039-41 hypothermia. Resuscitation 2004 Feb; 60 (2): 225-30
61. Boucher BA, Wood GC, Swanson JM. Pharmacokinetic 78. Roberts DJ, Goralski KB. A critical overview of the influ-
changes in critical illness. Crit Care Clin 2006 Apr; 22 (2): ence of inflammation and infection on P-glycoprotein
255-71, vi expression and activity in the brain. Expert Opin Drug
Metab Toxicol 2008 Oct; 4 (10): 1245-64
62. Butler JM, Begg EJ. Free drug metabolic clearance in el-
derly people. Clin Pharmacokinet 2008; 47 (5): 297-321 79. Roberts DJ, Goralski KB, Renton KW, et al. Effect
of acute inflammatory brain injury on accumulation of
63. Aaltonen L, Kanto J, Arola M, et al. Effect of age and
morphine and morphine 3- and 6-glucuronide in the
cardiopulmonary bypass on the pharmacokinetics of
human brain. Crit Care Med 2009 Oct; 37 (10): 2767-74
lorazepam. Acta Pharmacol et Toxicol 1982; 51: 126-31
80. Kollef MH, Levy NT, Ahrens TS, et al. The use of con-
64. Milne RW, Nation RL, Somogyi AA, et al. The influence tinuous i.v. sedation is associated with prolongation of
of renal function on the renal clearance of morphine and
mechanical ventilation. Chest 1998 Aug; 114 (2): 541-8
its glucuronide metabolites in intensive-care patients. Br J
Clin Pharmacol 1992 Jul; 34 (1): 53-9 81. Weinert CR, Calvin AD. Epidemiology of sedation and
sedation adequacy for mechanically ventilated patients in
65. Park GR, Miller E. What changes drug metabolism in cri- a medical and surgical intensive care unit. Crit Care Med
tically ill patients: III? Effect of pre-existing disease on the 2007 Feb; 35 (2): 393-401
metabolism of midazolam. Anaesthesia 1996 May; 51 (5):
431-4 82. Mehta S, Burry L, Fischer S, et al. Canadian survey of the use
of sedatives, analgesics, and neuromuscular blocking agents
66. Fukasawa T, Suzuki A, Otani K. Effects of genetic poly- in critically ill patients. Crit Care Med 2006 Feb; 34 (2): 374-80
morphism of cytochrome P450 enzymes on the pharma-
83. Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond
cokinetics of benzodiazepines. J Clin Pharm Ther 2007
Agitation-Sedation Scale: validity and reliability in adult
Aug; 32 (4): 333-41
intensive care unit patients. Am J Respir Crit Care Med
67. Hughes MA, Glass PS, Jacobs JR. Context-sensitive half- 2002 Nov 15; 166 (10): 1338-44
time in multicompartment pharmacokinetic models for
84. De JB, Cook D, Appere-De-Vecchi C, et al. Using and
intravenous anesthetic drugs. Anesthesiology 1992 Mar;
understanding sedation scoring systems: a systematic re-
76 (3): 334-41
view. Intensive Care Med 2000 Mar; 26 (3): 275-85
68. Albanese J, Martin C, Lacarelle B, et al. Pharmacokinetics
85. Khan BA, Guzman O, Campbell NL, et al. Comparison
of long-term propofol infusion used for sedation in ICU
and agreement between two sedation scales; Richmond
patients. Anesthesiology 1990 Aug; 73 (2): 214-7 Agitation Sedation Scale and Sedation Agitation Scale in
69. Park GR. Molecular mechanisms of drug metabolism in evaluating patients’ eligibility for delirium assessment in
the critically ill. Br J Anaesth 1996 Jul; 77 (1): 32-49 the intensive care unit. Chest. Epub 2012 Apr 26
70. Shelly MP, Mendel L, Park GR. Failure of critically ill 86. Ramsay MA, Savege TM, Simpson BR, et al. Controlled
patients to metabolise midazolam. Anaesthesia 1987 Jun; sedation with alphaxalone-alphadolone. Br Med J 1974
42 (6): 619-26 Jun 22; 2 (920): 656-9
71. Oldenhof H, de Jong M, Steenhoek A, et al. Clinical 87. Riker RR, Fraser GL, Cox PM. Continuous infusion of
pharmacokinetics of midazolam in intensive care patients, haloperidol controls agitation in critically ill patients. Crit
a wide interpatient variability? Clin Pharmacol Ther 1988 Care Med 1994 Mar; 22 (3): 433-40
Mar; 43 (3): 263-9 88. Ely EW, Truman B, Shintani A, et al. Monitoring sedation
72. Kharasch ED, Jubert C, Senn T, et al. Intraindividual status over time in ICU patients: reliability and validity of
variability in male hepatic CYP3A4 activity assessed by the Richmond Agitation-Sedation Scale (RASS). JAMA
alfentanil and midazolam clearance. J Clin Pharmacol 2003 Jun 11; 289 (22): 2983-91
1999 Jul; 39 (7): 664-9 89. Hardin KA, Seyal M, Stewart T, et al. Sleep in critically ill
73. Gorski JC, Jones DR, Haehner-Daniels BD, et al. The chemically paralyzed patients requiring mechanical ven-
contribution of intestinal and hepatic CYP3A to the tilation. Chest 2006 Jun; 129 (6): 1468-77

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1908 Roberts et al.

90. Crippen D. Role of bedside electroencephalography in the Ramsay and observer scales. Eur J Anaesthesiol 2006
adult intensive care unit during therapeutic neuromus- Aug; 23 (8): 649-53
cular blockade. Crit Care 1997; 1 (1): 15-24 108. Yamashita K, Terao Y, Inadomi C, et al. Age-dependent
91. Rampil IJ. A primer for EEG signal processing in an- relationship between Bispectral Index and sedation level.
esthesia. Anesthesiology 1998 Oct; 89 (4): 980-1002 J Clin Anesth 2008 Nov; 20 (7): 492-5
92. Johansen JW. Update on bispectral index monitoring. Best 109. Consales G, Chelazzi C, Rinaldi S, et al. Bispectral Index
Pract Res Clin Anaesthesiol 2006 Mar; 20 (1): 81-99 compared to Ramsay score for sedation monitoring in
93. Avidan MS, Zhang L, Burnside BA, et al. Anesthesia intensive care units. Minerva Anestesiol 2006 May; 72 (5):
awareness and the bispectral index. N Engl J Med 2008 329-36
Mar 13; 358 (11): 1097-108 110. Haenggi M, Ypparila-Wolters H, Bieri C, et al. Entropy
94. Liu J, Singh H, White PF. Electroencephalographic bis- and bispectral index for assessment of sedation, analgesia
pectral index correlates with intraoperative recall and and the effects of unpleasant stimuli in critically ill patients:
depth of propofol-induced sedation. Anesth Analg 1997 an observational study. Crit Care 2008; 12 (5): R119
Jan; 84 (1): 185-9 111. Arbour R, Waterhouse J, Seckel MA, et al. Correlation
95. Ekman A, Lindholm ML, Lennmarken C, et al. Reduction between the Sedation-Agitation Scale and the Bispectral
in the incidence of awareness using BIS monitoring. Acta Index in ventilated patients in the intensive care unit.
Anaesthesiol Scand 2004 Jan; 48 (1): 20-6 Heart Lung 2009 Jul-Aug; 38 (4): 336-45
96. Leslie K, Myles PS, Forbes A, et al. The effect of bispectral 112. LeBlanc JM, Dasta JF, Pruchnicki MC, et al. Bispectral
index monitoring on long-term survival in the B-aware index values, sedation-agitation scores, and plasma Lora-
trial. Anesth Analg 2010 Mar 1; 110 (3): 816-22 zepam concentrations in critically ill surgical patients. Am
97. Karamchandani K, Rewari V, Trikha A, et al. Bispectral J Crit Care 2012 Mar; 21 (2): 99-105
index correlates well with Richmond agitation sedation 113. Vivien B, Di MS, Ouattara A, et al. Overestimation of
scale in mechanically ventilated critically ill patients. Bispectral Index in sedated intensive care unit patients
J Anesth 2010 Jun; 24 (3): 394-8 revealed by administration of muscle relaxant. Anesthe-
98. Deogaonkar A, Gupta R, DeGeorgia M, et al. Bispectral siology 2003 Jul; 99 (1): 9-17
Index monitoring correlates with sedation scales in brain- 114. Nasraway Jr SA, Wu EC, Kelleher RM, et al. How reliable
injured patients. Crit Care Med 2004 Dec; 32 (12): 2403-6 is the Bispectral Index in critically ill patients? A pro-
99. Turkmen A, Altan A, Turgut N, et al. The correlation spective, comparative, single-blinded observer study. Crit
between the Richmond agitation-sedation scale and bis- Care Med 2002 Jul; 30 (7): 1483-7
pectral index during dexmedetomidine sedation. Eur J 115. Sackey PV, Radell PJ, Granath F, et al. Bispectral index as
Anaesthesiol 2006 Apr; 23 (4): 300-4 a predictor of sedation depth during isoflurane or mid-
100. Ogilvie MP, Pereira BM, Ryan ML, et al. Bispectral index azolam sedation in ICU patients. Anaesth Intensive Care
to monitor propofol sedation in trauma patients. J Trau- 2007 Jun; 35 (3): 348-56
ma 2011 Nov; 71 (5): 1415-21 116. Weatherburn C, Endacott R, Tynan P, et al. The impact of
101. Mondello E, Siliotti R, Noto G, et al. Bispectral Index in Bispectral Index monitoring on sedation administration
ICU: correlation with Ramsay Score on assessment of in mechanically ventilated patients. Anaesth Intensive
sedation level. J Clin Monit Comput 2002 Jul; 17 (5): 271-7 Care 2007 Apr; 35 (2): 204-8
102. Hernandez-Gancedo C, Pestana D, Perez-Chrzanowska H, 117. Inoue S, Kawaguchi M, Sasaoka N, et al. Effects of neuro-
et al. Comparing entropy and the Bispectral Index with muscular block on systemic and cerebral hemodynamics
the Ramsay score in sedated ICU patients. J Clin Monit and bispectral index during moderate or deep sedation in
Comput 2007 Oct; 21 (5): 295-302 critically ill patients. Intensive Care Med 2006 Mar; 32 (3):
103. Olson DM, Thoyre SM, Peterson ED, et al. A randomized 391-7
evaluation of bispectral index-augmented sedation as- 118. Dahaba AA, Mattweber M, Fuchs A, et al. The effect of
sessment in neurological patients. Neurocrit Care 2009; 11 different stages of neuromuscular block on the bispectral
(1): 20-7 index and the bispectral index-XP under remifentanil/
104. Lu CH, Ou-Yang HY, Man KM, et al. Relative reliability propofol anesthesia. Anesth Analg 2004 Sep; 99 (3): 781-7
of the auditory evoked potential and Bispectral Index for 119. Trouiller P, Fangio P, Paugam-Burtz C, et al. Frequency
monitoring sedation level in surgical intensive care and clinical impact of preserved Bispectral Index activity
patients. Anaesth Intensive Care 2008 Jul; 36 (4): 553-9 during deep sedation in mechanically ventilated ICU
105. Lu CH, Man KM, Ou-Yang HY, et al. Composite auditory patients. Intensive Care Med 2009 Dec; 35 (12): 2096-104
evoked potential index versus bispectral index to estimate 120. LeBlanc JM, Dasta JF, Kane-Gill SL. Role of the bispec-
the level of sedation in paralyzed critically ill patients: a tral index in sedation monitoring in the ICU. Ann
prospective observational study. Anesth Analg 2008 Oct; Pharmacother 2006 Mar; 40 (3): 490-500
107 (4): 1290-4 121. Walsh TS, Ramsay P, Lapinlampi TP, et al. An assessment
106. Adesanya AO, Rosero E, Wyrick C, et al. Assessing the of the validity of spectral entropy as a measure of sedation
predictive value of the bispectral index vs patient state state in mechanically ventilated critically ill patients. In-
index on clinical assessment of sedation in postoperative tensive Care Med 2008 Feb; 34 (2): 308-15
cardiac surgery patients. J Crit Care 2009 Sep; 24 (3): 322-8 122. Springman SR, Andrei AC, Willmann K, et al. A comparison
107. Hernandez-Gancedo C, Pestana D, Pena N, et al. Mon- of SNAP II and Bispectral Index monitoring in patients
itoring sedation in critically ill patients: bispectral index, undergoing sedation. Anaesthesia 2010 Aug; 65 (8): 815-9

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1909

123. Weber F, Steinberger M, Ritzka M, et al. Measuring depth 139. Meissner W, Dohrn B, Reinhart K. Enteral naloxone re-
of sedation in intensive care patients with the electro- duces gastric tube reflux and frequency of pneumonia in
encephalographic Narcotrend Index. Eur J Anaesthesiol critical care patients during opioid analgesia. Crit Care
2008 Feb; 25 (2): 123-8 Med 2003 Mar; 31 (3): 776-80
124. Riker RR, Fraser GL. Adverse events associated with sed- 140. Collop NA, Salas RE, Delayo M, et al. Normal sleep and
atives, analgesics, and other drugs that provide patient circadian processes. Crit Care Clin 2008 Jul; 24 (3): 449-60, v
comfort in the intensive care unit. Pharmacotherapy 2005 141. Weinhouse GL, Watson PL. Sedation and sleep dis-
May; 25 (5 Pt 2): 8S-18S turbances in the ICU. Anesthesiol Clin 2011 Dec; 29 (4):
125. Ouimet S, Kavanagh BP, Gottfried SB, et al. Incidence, 675-85
risk factors and consequences of ICU delirium. Intensive 142. Salas RE, Gamaldo CE. Adverse effects of sleep depriva-
Care Med 2007 Jan; 33 (1): 66-73 tion in the ICU. Crit Care Clin 2008 Jul; 24 (3): 461-76, v-vi
126. Van Rompaey B, Elseviers MM, Schuurmans MJ, et al. 143. Little A, Ethier C, Ayas N, et al. A patient survey of sleep
Risk factors for delirium in intensive care patients: a quality in the intensive care unit. Minerva Anestesiol 2012
prospective cohort study. Crit Care 2009; 13 (3): R77 Apr; 78 (4): 406-14
127. Ely EW, Inouye SK, Bernard GR, et al. Delirium in me- 144. Freedman NS, Kotzer N, Schwab RJ. Patient perception of
chanically ventilated patients: validity and reliability of sleep quality and etiology of sleep disruption in the in-
the confusion assessment method for the intensive care tensive care unit. Am J Respir Crit Care Med 1999 Apr;
unit (CAM-ICU). JAMA 2001 Dec 5; 286 (21): 2703-10 159 (4 Pt 1): 1155-62
128. Ely EW, Shintani A, Truman B, et al. Delirium as a pre- 145. Gabor JY, Cooper AB, Crombach SA, et al. Contribution
dictor of mortality in mechanically ventilated patients in of the intensive care unit environment to sleep disruption
the intensive care unit. JAMA 2004 Apr 14; 291 (14): in mechanically ventilated patients and healthy subjects.
1753-62 Am J Respir Crit Care Med 2003 Mar 1; 167 (5): 708-15
129. Shehabi Y, Riker RR, Bokesch PM, et al. Delirium dura- 146. Meyer TJ, Eveloff SE, Bauer MS, et al. Adverse environ-
tion and mortality in lightly sedated, mechanically venti- mental conditions in the respiratory and medical ICU
lated intensive care patients. Crit Care Med 2010 Dec; 38 settings. Chest 1994 Apr; 105 (4): 1211-6
(12): 2311-8
147. Cooper AB, Thornley KS, Young GB, et al. Sleep in criti-
130. Milbrandt EB, Deppen S, Harrison PL, et al. Costs asso- cally ill patients requiring mechanical ventilation. Chest
ciated with delirium in mechanically ventilated patients. 2000 Mar; 117 (3): 809-18
Crit Care Med 2004 Apr; 32 (4): 955-62
148. Sanders RD, Maze M. Contribution of sedative-hypnotic
131. Girard TD, Jackson JC, Pandharipande PP, et al. Delirium agents to delirium via modulation of the sleep pathway.
as a predictor of long-term cognitive impairment in sur- Can J Anaesth 2011 Feb; 58 (2): 149-56
vivors of critical illness. Crit Care Med 2010 Jul; 38 (7):
1513-20 149. Spiegel K, Sheridan JF, Van CE. Effect of sleep deprivation
on response to immunization. JAMA 2002 Sep 25; 288
132. Mazoit JX, Butscher K, Samii K. Morphine in post- (12): 1471-2
operative patients: pharmacokinetics and pharmaco-
dynamics of metabolites. Anesth Analg 2007 Jul; 105 (1): 150. Dinges DF, Douglas SD, Hamarman S, et al. Sleep depri-
70-8 vation and human immune function. Adv Neuroimmunol
1995; 5 (2): 97-110
133. Bauer TM, Ritz R, Haberthur C, et al. Prolonged sedation
due to accumulation of conjugated metabolites of mid- 151. Chen HI, Tang YR. Sleep loss impairs inspiratory muscle
azolam. Lancet 1995 Jul 15; 346 (8968): 145-7 endurance. Am Rev Respir Dis 1989 Oct; 140 (4): 907-9
134. Kharasch ED, Thummel KE. Human alfentanil metabo- 152. Gallicchio L, Kalesan B. Sleep duration and mortality: a
lism by cytochrome P450 3A3/4: an explanation for the systematic review and meta-analysis. J Sleep Res 2009
interindividual variability in alfentanil clearance? Anesth Jun; 18 (2): 148-58
Analg 1993 May; 76 (5): 1033-9 153. Bixler EO, Kales JD, Kales A, et al. Rebound insomnia and
135. Kharasch ED, Walker A, Isoherranen N, et al. Influence of elimination half-life: assessment of individual subject re-
CYP3A5 genotype on the pharmacokinetics and pharma- sponse. J Clin Pharmacol 1985 Mar; 25 (2): 115-24
codynamics of the cytochrome P4503A probes alfentanil 154. Cammarano WB, Pittet JF, Weitz S, et al. Acute with-
and midazolam. Clin Pharmacol Ther 2007 Oct; 82 (4): drawal syndrome related to the administration of an-
410-26 algesic and sedative medications in adult intensive care
136. Lin YS, Dowling AL, Quigley SD, et al. Co-regulation of unit patients. Crit Care Med 1998 Apr; 26 (4): 676-84
CYP3A4 and CYP3A5 and contribution to hepatic and 155. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epide-
intestinal midazolam metabolism. Mol Pharmacol 2002 miology of severe sepsis in the United States: analysis of
Jul; 62 (1): 162-72 incidence, outcome, and associated costs of care. Crit
137. Chua MV, Tsueda K, Doufas AG. Midazolam causes less Care Med 2001 Jul; 29 (7): 1303-10
sedation in volunteers with red hair. Can J Anaesth 2004 156. Vincent JL, Sakr Y, Sprung CL, et al. Sepsis in European
Jan; 51 (1): 25-30 intensive care units: results of the SOAP study. Crit Care
138. Morris AC, Hay AW, Swann DG, et al. Reducing venti- Med 2006 Feb; 34 (2): 344-53
lator-associated pneumonia in intensive care: impact of 157. Vincent JL, Rello J, Marshall J, et al. International study of
implementing a care bundle. Crit Care Med 2011 Oct; 39 the prevalence and outcomes of infection in intensive care
(10): 2218-24 units. JAMA 2009 Dec 2; 302 (21): 2323-9

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1910 Roberts et al.

158. Pandharipande PP, Sanders RD, Girard TD, et al. Effect eclampsia patients in the intensive care unit. J Crit Care
of dexmedetomidine versus lorazepam on outcome in 2009 Dec; 24 (4): 551-5
patients with sepsis: an a priori-designed analysis of the 174. Ruokonen E, Parviainen I, Jakob SM, et al. Dexmede-
MENDS randomized controlled trial. Crit Care 2010; 14 tomidine versus propofol/midazolam for long-term seda-
(2): R38 tion during mechanical ventilation. Intensive Care Med
159. Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine 2009 Feb; 35 (2): 282-90
vs midazolam for sedation of critically ill patients: a ran- 175. Shehabi Y, Grant P, Wolfenden H, et al. Prevalence of
domized trial. JAMA 2009 Feb 4; 301 (5): 489-99 delirium with dexmedetomidine compared with morphine
160. Memis D, Hekimoglu S, Vatan I, et al. Effects of mid- based therapy after cardiac surgery: a randomized con-
azolam and dexmedetomidine on inflammatory responses trolled trial (DEXmedetomidine COmpared to Morphine-
and gastric intramucosal pH to sepsis, in critically ill DEXCOM Study). Anesthesiology 2009 Nov; 111 (5):
patients. Br J Anaesth 2007 Apr; 98 (4): 550-2 1075-84
161. Nseir S, Makris D, Mathieu D, et al. Intensive care unit- 176. Maldonado JR, Wysong A, van der Starre PJ, et al.
acquired infection as a side effect of sedation. Crit Care Dexmedetomidine and the reduction of postoperative
2010; 14 (2): R30 delirium after cardiac surgery. Psychosomatics 2009 May-
162. Sanders RD, Hussell T, Maze M. Sedation and immuno- Jun; 50 (3): 206-17
modulation. Anesthesiol Clin 2011 Dec; 29 (4): 687-706 177. Sackey PV, Martling CR, Carlsward C, et al. Short- and
163. Venn RM, Grounds RM. Comparison between dexmede- long-term follow-up of intensive care unit patients after
tomidine and propofol for sedation in the intensive care sedation with isoflurane and midazolam: a pilot study.
unit: patient and clinician perceptions. Br J Anaesth 2001 Crit Care Med 2008 Mar; 36 (3): 801-6
Nov; 87 (5): 684-90 178. Huey-Ling L, Chun-Che S, Jen-Jen T, et al. Comparison of
164. Ma D, Hossain M, Rajakumaraswamy N, et al. the effect of protocol-directed sedation with propofol vs.
Dexmedetomidine produces its neuroprotective effect via midazolam by nurses in intensive care: efficacy, haemo-
the alpha 2A-adrenoceptor subtype. Eur J Pharmacol dynamic stability and patient satisfaction. J Clin Nurs
2004 Oct 11; 502 (1-2): 87-97 2008 Jun; 17 (11): 1510-7
165. Jakob SM, Ruokonen E, Grounds RM, et al. Dexmede- 179. Pandharipande PP, Pun BT, Herr DL, et al. Effect of se-
tomidine vs midazolam or propofol for sedation during dation with dexmedetomidine vs lorazepam on acute
prolonged mechanical ventilation: two randomized con- brain dysfunction in mechanically ventilated patients: the
trolled trials. JAMA 2012 Mar 21; 307 (11): 1151-60 MENDS randomized controlled trial. JAMA 2007 Dec
166. Bjelland TW, Dale O, Kaisen K, et al. Propofol and remi- 12; 298 (22): 2644-53
fentanil versus midazolam and fentanyl for sedation during 180. Ghori KA, Harmon DC, Elashaal A, et al. Effect of mid-
therapeutic hypothermia after cardiac arrest: a randomised azolam versus propofol sedation on markers of neuro-
trial. Intensive Care Med 2012 Jun; 38 (6): 959-67 logical injury and outcome after isolated severe head injury:
167. Hellstrom J, Owall A, Bergstrom J, et al. Cardiac outcome a pilot study. Crit Care Resusc 2007 Jun; 9 (2): 166-71
after sevoflurane versus propofol sedation following cor- 181. Richman PS, Baram D, Varela M, et al. Sedation during
onary bypass surgery: a pilot study. Acta Anaesthesiol mechanical ventilation: a trial of benzodiazepine and
Scand 2011 Apr; 55 (4): 460-7 opiate in combination. Crit Care Med 2006 May; 34 (5):
168. Hellstrom J, Owall A, Sackey PV. Wake-up times following 1395-401
sedation with sevoflurane versus propofol after cardiac 182. Hsiao PC, Tang YY, Liaw WJ, et al. Postoperative seda-
surgery. Scand Cardiovasc J. Epub 2012 Mar 29 tion after major surgery with midazolam or propofol in
169. Candiotti KA, Gan TJ, Young C, et al. A randomized, the ICU: effects on amnesia and anxiety. Acta Anaes-
open-label study of the safety and tolerability of fospro- thesiol Taiwan 2006 Jun; 44 (2): 93-9
pofol for patients requiring intubation and mechanical 183. Ibrahim MG, Bellomo R, Hart GK, et al. A double-blind
ventilation in the intensive care unit. Anesth Analg 2011 placebo-controlled randomised pilot study of nocturnal
Sep; 113 (3): 550-6 melatonin in tracheostomised patients. Crit Care Resusc
170. Mesnil M, Capdevila X, Bringuier S, et al. Long-term se- 2006 Sep; 8 (3): 187-91
dation in intensive care unit: a randomized comparison 184. Breen D, Karabinis A, Malbrain M, et al. Decreased
between inhaled sevoflurane and intravenous propofol or duration of mechanical ventilation when comparing
midazolam. Intensive Care Med 2011 Jun; 37 (6): 933-41 analgesia-based sedation using remifentanil with standard
171. Mirski MA, Lewin III JJ, Ledroux S, et al. Cognitive im- hypnotic-based sedation for up to 10 days in intensive care
provement during continuous sedation in critically ill, unit patients: a randomised trial [ISRCTN47583497]. Crit
awake and responsive patients: the Acute Neurological Care 2005; 9 (3): R200-10
ICU Sedation Trial (ANIST). Intensive Care Med 2010 185. Corbett SM, Rebuck JA, Greene CM, et al. Dexmede-
Sep; 36 (9): 1505-13 tomidine does not improve patient satisfaction when
172. Tasdogan M, Memis D, Sut N, et al. Results of a pilot compared with propofol during mechanical ventilation.
study on the effects of propofol and dexmedetomidine on Crit Care Med 2005 May; 33 (5): 940-5
inflammatory responses and intraabdominal pressure in 186. Bourgoin A, Albanese J, Leone M, et al. Effects of sufent-
severe sepsis. J Clin Anesth 2009 Sep; 21 (6): 394-400 anil or ketamine administered in target-controlled infusion
173. Esmaoglu A, Ulgey A, Akin A, et al. Comparison between on the cerebral hemodynamics of severely brain-injured
dexmedetomidine and midazolam for sedation of patients. Crit Care Med 2005 May; 33 (5): 1109-13

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1911

187. Karabinis A, Mandragos K, Stergiopoulos S, et al. Safety containing EDTA versus other sedatives in critically ill
and efficacy of analgesia-based sedation with remifentanil patients. Intensive Care Med 2000; 26 Suppl. 4: S413-21
versus standard hypnotic-based regimens in intensive care 202. Sandiumenge CA, Sanchez-Izquierdo Riera JA, Toral VD,
unit patients with brain injuries: a randomised, controlled et al. Midazolam and 2% propofol in long-term seda-
trial [ISRCTN50308308]. Crit Care 2004; 8 (4): R268-80 tion of traumatized critically ill patients: efficacy and
188. Martin E, Ramsay G, Mantz J, et al. The role of the alpha2- safety comparison. Crit Care Med 2000 Nov; 28 (11):
adrenoceptor agonist dexmedetomidine in postsurgical 3612-9
sedation in the intensive care unit. J Intensive Care Med 203. McCollam JS, O’Neil MG, Norcross ED, et al. Continuous
2003 Jan-Feb; 18 (1): 29-41 infusions of lorazepam, midazolam, and propofol for se-
189. Herr DL, Sum-Ping ST, England M. ICU sedation after dation of the critically ill surgery trauma patient: a pro-
coronary artery bypass graft surgery: dexmedetomidine- spective, randomized comparison. Crit Care Med 1999
based versus propofol-based sedation regimens. J Cardi- Nov; 27 (11): 2454-8
othorac Vasc Anesth 2003 Oct; 17 (5): 576-84 204. Swart EL, van Schijndel RJ, van Loenen AC, et al. Con-
190. Saito M, Terao Y, Fukusaki M, et al. Sequential use of tinuous infusion of lorazepam versus medazolam in
midazolam and propofol for long-term sedation in post- patients in the intensive care unit: sedation with loraze-
operative mechanically ventilated patients. Anesth Analg pam is easier to manage and is more cost-effective. Crit
2003 Mar; 96 (3): 834-8 Care Med 1999 Aug; 27 (8): 1461-5
191. Bourgoin A, Albanese J, Wereszczynski N, et al. Safety of 205. Venn RM, Bradshaw CJ, Spencer R, et al. Preliminary UK
sedation with ketamine in severe head injury patients: experience of dexmedetomidine, a novel agent for post-
comparison with sufentanil. Crit Care Med 2003 Mar; 31 operative sedation in the intensive care unit. Anaesthesia
(3): 711-7 1999 Dec; 54 (12): 1136-42
192. Triltsch AE, Welte M, von Homeyer P, et al. Bispectral 206. Kelly DF, Goodale DB, Williams J, et al. Propofol in the
index-guided sedation with dexmedetomidine in intensive treatment of moderate and severe head injury: a ran-
care: a prospective, randomized, double blind, placebo- domized, prospective double-blinded pilot trial. J Neuro-
controlled phase II study. Crit Care Med 2002 May; 30 surg 1999 Jun; 90 (6): 1042-52
(5): 1007-14 207. Carrasco G, Cabré L, Sobrepere G, et al. Synergistic se-
193. Walder B, Borgeat A, Suter PM, et al. Propofol and dation with propofol and midazolam in intensive care
midazolam versus propofol alone for sedation following patients after coronary artery bypass grafting. Crit Care
coronary artery bypass grafting: a randomized, placebo- Med 1998 May; 26 (5): 844-51
controlled trial. Anaesth Intensive Care 2002 Apr; 30 (2): 208. Sanchez-Izquierdo-Riera JA, Caballero-Cubedo RE, Per-
171-8 ez-Vela JL, et al. Propofol versus midazolam: safety and
194. Venn RM, Bryant A, Hall GM, et al. Effects of dexmede- efficacy for sedating the severe trauma patient. Anesth
tomidine on adrenocortical function, and the cardio- Analg 1998 Jun; 86 (6): 1219-24
vascular, endocrine and inflammatory responses in post- 209. Weinbroum AA, Halpern P, Rudick V, et al. Midazolam
operative patients needing sedation in the intensive care versus propofol for long-term sedation in the ICU: a
unit. Br J Anaesth 2001 May; 86 (5): 650-6 randomized prospective comparison. Intensive Care Med
195. Barr J, Zomorodi K, Bertaccini EJ, et al. A double-blind, 1997 Dec; 23 (12): 1258-63
randomized comparison of i.v. lorazepam versus mid- 210. Searle NR, Cote S, Taillefer J, et al. Propofol or midazolam
azolam for sedation of ICU patients via a pharmacologic for sedation and early extubation following cardiac sur-
model. Anesthesiology 2001 Aug; 95 (2): 286-98 gery. Can J Anaesth 1997 Jun; 44 (6): 629-35
196. Hall RI, Sandham D, Cardinal P, et al. Propofol vs mid- 211. Plunkett JJ, Reeves JD, Ngo L, et al. Urine and plasma
azolam for ICU sedation: a Canadian multicenter ran- catecholamine and cortisol concentrations after myo-
domized trial. Chest 2001 Apr; 119 (4): 1151-9 cardial revascularization: modulation by continuous
197. Kress JP, Pohlman AS, Hall JB. Effects of sedative inter- sedation. Multicenter Study of Perioperative Ischemia
ruption in critically ill, mechanically ventilated patients (McSPI) Research Group, and the Ischemia Research and
receiving midazolam or propofol. J Clin Outcomes Man- Education Foundation (IREF). Anesthesiology 1997; 86
age 2001; 8 (2): 33-9 (4): 785-96
198. Herr DL, Kelly K, Hall JB, et al. Safety and efficacy of 212. Barrientos-Vega R, Mar S, Morales-Garcia C, et al. Pro-
propofol with EDTA when used for sedation of surgical longed sedation of critically ill patients with midazolam or
intensive care unit patients. Intensive Care Med 2000; 26 propofol: impact on weaning and costs. Crit Care Med
Suppl. 4: S452-62 1997 Jan; 25 (1): 33-40
199. Barr J, Zaloga GP, Haupt MT, et al. Cation metabolism 213. Manley NM, Fitzpatrick RW, Long T, et al. A cost analysis
during propofol sedation with and without EDTA in of alfentanil+propofol vs morphine+midazolam for the
patients with impaired renal function. Intensive Care Med sedation of critically ill patients. Pharmacoeconomics
2000; 26 Suppl. 4: S433-42 1997 Aug; 12 (2 Pt 2): 247-55
200. Abraham E, Papadakos PJ, Tharratt RS, et al. Effects of 214. Treggiari-Venzi M, Borgeat A, Fuchs-Buder T, et al.
propofol containing EDTA on mineral metabolism in Overnight sedation with midazolam or propofol in the
medical ICU patients with pulmonary dysfunction. In- ICU: effects on sleep quality, anxiety and depression. In-
tensive Care Med 2000; 26 Suppl. 4: S422-32 tensive Care Med 1996 Nov; 22 (11): 1186-90
201. Higgins TL, Murray M, Kett DH, et al. Trace element 215. Chamorro C, de Latorre FJ, Montero A, et al. Compara-
homeostasis during continuous sedation with propofol tive study of propofol versus midazolam in the sedation of

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1912 Roberts et al.

critically ill patients: results of a prospective, randomized, parative study with lytic mixture. Anaesthesiol Reanim
multicenter trial. Crit Care Med 1996 Jun; 24 (6): 932-9 1992; 17 (2): 77-6
216. Wahr JA, Plunkett JJ, Ramsay JG, et al. Cardiovascular 231. Spencer EM, Willatts SM. Isoflurane for prolonged seda-
responses during sedation after coronary revasculariza- tion in the intensive care unit; efficacy and safety. In-
tion: incidence of myocardial ischemia and hemodynamic tensive Care Med 1992; 18 (7): 415-21
episodes with propofol versus midazolam. Institutions of 232. Beyer R, Seyde WC. Propofol versus midazolam: long-
the McSPI Research Group. Anesthesiology 1996 Jun; 84 term sedation in the intensive care unit. Anaesthesist 1992;
(6): 1350-60 41 (6): 335-41
217. Kolenda H, Gremmelt A, Rading S, et al. Ketamine for 233. Clarke TNS. Propofol compared with midazolam for se-
analgosedative therapy in intensive care treatment of dation following prolonged neurosurgery. J Drug Devel-
head-injured patients. Acta Neurochir (Wien) 1996; 138 opment 1991; 4 Suppl. 3: 108-9
(10): 1193-9
234. Boyle WA, Shear JM, White PF, et al. Long-term sedative
218. Cernaianu AC, Del Rossi AJ, Flum DR, et al. Lorazepam infusion in the intensive care unit: propofol versus mid-
and midazolam in the intensive care unit: a randomized, azolam. J Drug Development 1991; 4 Suppl. 3: 43-5
prospective, multicenter study of hemodynamics, oxygen
transport, efficacy, and cost. Crit Care Med 1996 Feb; 24 235. Beyer R, Seyde WC. Long-term sedation (24h) in the in-
(2): 222-8 tensive care unit: a comparison of propofol and mid-
azolam. J Drug Development 1991; 4 Suppl. 3: 67-8
219. Kress JP, O’Connor MF, Pohlman AS, et al. Sedation of
critically ill patients during mechanical ventilation: a 236. Degauque C, Dupuis A. A study to compare the use of
comparison of propofol and midazolam. Am J Respir Crit propofol and midazolam for the sedation of patients with
Care Med 1996 Mar; 153 (3): 1012-8 acute respiratory failure. J Drug Development 1991; 4
Suppl. 3: 95-7
220. McArthur CJ, Gin T, McLaren IM, et al. Gastric emptying
following brain injury: effects of choice of sedation and 237. Wolfs C, Kimbimbi P, Colin FNF, et al. A comparison of
intracranial pressure. Intensive Care Med 1995 Jul; 21 (7): propofol/fentanyl and midazolam/fentanyl for ICU seda-
573-6 tion after abdominal surgery. J Drug Development 1991;
4 Suppl. 3: 69-71
221. Ronan KP, Gallagher TJ, George B, et al. Comparison of
propofol and midazolam for sedation in intensive care 238. Adams HA, Claussen E, Gebhardt B, et al. The use of
unit patients. Crit Care Med 1995 Feb; 23 (2): 286-93 ketamine and midazolam for analgesia and sedation in
ventilated patients subject to obligatory treatment with
222. Levy ML, Aranda M, Zelman V, et al. Propylene glycol
catecholamines. Anaesthesist 1991 Apr; 40 (4): 238-44
toxicity following continuous etomidate infusion for the
control of refractory cerebral edema. Neurosurgery 1995 239. McMurray TJ, Collier PS, Carson IW, et al. Propofol sed-
Aug; 37 (2): 363-9 ation after open heart surgery: a clinical and pharma-
cokinetic study. Anaesthesia 1990 Apr; 45 (4): 322-6
223. Higgins TL, Yared JP, Estafanous FG, et al. Propofol
versus midazolam for intensive care unit sedation after 240. Snellen F, Lauwers P, Demeyere R, et al. The use of mid-
coronary artery bypass grafting. Crit Care Med 1994 Sep; azolam versus propofol for short-term sedation following
22 (9): 1415-23 coronary artery bypass grafting. Intensive Care Med
1990; 16 (5): 312-6
224. Pohlman AS, Simpson KP, Hall JB. Continuous intravenous
infusions of lorazepam versus midazolam for sedation 241. Harris CE, Grounds RM, Murray AM, et al. Propofol for
during mechanical ventilatory support: a prospective, ran- long-term sedation in the intensive care unit: a compar-
domized study. Crit Care Med 1994 Aug; 22 (8): 1241-7 ison with papaveretum and midazolam. Anaesthesia 1990
225. Costa J, Cabre L, Molina R, et al. Cost of ICU sedation: May; 45 (5): 366-72
comparison of empirical and controlled sedation meth- 242. Gottardis M, Khunl-Brady KS, Koller W, et al. Effect of
ods. Clin Intensive Care 1994; 5 (5 Suppl.): 17-21 prolonged sedation with propofol on serum triglyceride
226. Carrasco G, Molina R, Costa J, et al. Propofol vs mid- and cholesterol concentrations. Br J Anaesth 1989 Apr; 62
azolam in short-, medium-, and long-term sedation of (4): 393-6
critically ill patients: a cost-benefit analysis. Chest 1993 243. Kong KL, Willatts SM, Prys-Roberts C. Isoflurane com-
Feb; 103 (2): 557-64 pared with midazolam for sedation in the intensive care
227. Roekaerts PM, Huygen FJ, de Lange S. Infusion of pro- unit. BMJ 1989 May 13; 298 (6683): 1277-80
pofol versus midazolam for sedation in the intensive care 244. Boeke A, Lauwers J, Schurink G. A pilot study to compare
unit following coronary artery surgery. J Cardiothorac the use of propofol and midazolam for long-term seda-
Vasc Anesth 1993 Apr; 7 (2): 142-7 tion. J Drug Development 1989; 2 Suppl. 2: 71-2
228. Boyd O, Mackay CJ, Rushmer F, et al. Propofol or mid- 245. Aitkenhead AR, Pepperman ML, Willatts SM, et al.
azolam for short-term alterations in sedation. Can J An- Comparison of propofol and midazolam for sedation in
aesth 1993 Dec; 40 (12): 1142-7 critically ill patients. Lancet 1989 Sep 23; 2 (8665): 704-9
229. Chaudhri S, Kenny GN. Sedation after cardiac bypass 246. Ledingham IM, Bion JF, Newman LH, et al. Mortality and
surgery: comparison of propofol and midazolam in the morbidity amongst sedated intensive care patients. Re-
presence of a computerized closed loop arterial pressure suscitation 1988; 16 Suppl.: S69-77
controller. Br J Anaesth 1992 Jan; 68 (1): 98-9 247. Adams HA, Biscoping J, Russ W, et al. Sedative-analgesic
230. Heinrichs W, Tzanova I, Brost F, et al. Propofol for seda- medication in intensive care patients needing ventilator
tion during postoperative mechanical ventilation: com- treatment. Anaesthesist 1988 Apr; 37 (4): 268-76

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1913

248. Grounds RM, Lalor JM, Lumley J, et al. Propofol infusion 266. Ahonen J, Olkkola KT, Takala A, et al. Interaction between
for sedation in the intensive care unit: preliminary report. fluconazole and midazolam in intensive care patients. Acta
Br Med J (Clin Res Ed) 1987 Feb 14; 294 (6569): 397-400 Anaesthesiol Scand 1999 May; 43 (5): 509-14
249. Dearden NM, McDowall DG. Comparison of etomidate and 267. Inomata S, Nagashima A, Itagaki F, et al. CYP2C19 geno-
althesin in the reduction of increased intracranial pressure type affects diazepam pharmacokinetics and emergence
after head injury. Br J Anaesth 1985 Apr; 57 (4): 361-8 from general anesthesia. Clin Pharmacol Ther 2005 Dec;
250. White PF, Schlobohm RM, Pitts LH, et al. A randomized 78 (6): 647-55
study of drugs for preventing increases in intracranial 268. Vellema J, Hunfeld NG, Van den Akker HE, et al. Avoid-
pressure during endotracheal suctioning. Anesthesiology ing crystallization of lorazepam during infusion. Eur J
1982 Sep; 57 (3): 242-4 Pharm Sci 2011 Dec 18; 44 (5): 621-6
251. Walder B, Elia N, Henzi I, et al. A lack of evidence of 269. Horinek EL, Kiser TH, Fish DN, et al. Propylene glycol
superiority of propofol versus midazolam for sedation in accumulation in critically ill patients receiving continuous
mechanically ventilated critically ill patients: a qualitative intravenous lorazepam infusions. Ann Pharmacother
and quantitative systematic review. Anesth Analg 2001 2009 Dec; 43 (12): 1964-71
Apr; 92 (4): 975-83 270. Yaucher NE, Fish JT, Smith HW, et al. Propylene glycol-
252. Magarey JM. Propofol or midazolam: which is best for the associated renal toxicity from lorazepam infusion.
sedation of adult ventilated patients in intensive care units? Pharmacotherapy 2003 Sep; 23 (9): 1094-9
A systematic review. Aust Crit Care 2001; 14 (4): 147-54 271. Olsen RW, Li GD. GABA(A) receptors as molecular tar-
253. Ho KM, Ng JY. The use of propofol for medium and long- gets of general anesthetics: identification of binding sites
term sedation in critically ill adult patients: a meta- provides clues to allosteric modulation. Can J Anaesth
analysis. Intensive Care Med 2008 Nov; 34 (11): 1969-79 2011 Feb; 58 (2): 206-15
254. Ghoneim MM, Mewaldt SP. Benzodiazepines and human 272. Gommers D, Bakker J. Medications for analgesia and se-
memory: a review. Anesthesiology 1990 May; 72 (5): 926-38 dation in the intensive care unit: an overview. Crit Care
255. Roberts DJ, Zygun D. Comparative efficacy and safety of 2008; 12 Suppl. 3: S4
sedative agents in severe traumatic brain injury. In: Vin- 273. Veselis RA, Reinsel RA, Wronski M, et al. EEG and
cent J-L, editor. Annual update in intensive care and memory effects of low-dose infusions of propofol. Br J
emergency medicine 2012. Berlin Heidelberg: Springer- Anaesth 1992 Sep; 69 (3): 246-54
Verlag, 2012: 771-82 274. Rossetti AO, Lowenstein DH. Management of refractory
256. Gilliland HE, Prasad BK, Mirakhur RK, et al. An in- status epilepticus in adults: still more questions than an-
vestigation of the potential morphine sparing effect of swers. Lancet Neurol 2011 Oct; 10 (10): 922-30
midazolam. Anaesthesia 1996 Sep; 51 (9): 808-11 275. Marik PE, Varon J. The management of status epilepticus.
257. Andersin R. Solubility and acid-base behaviour of mid- Chest 2004 Aug; 126 (2): 582-91
azolam in media of different pH, studied by ultraviolet 276. Mirenda J, Broyles G. Propofol as used for sedation in the
spectrophotometry with multicomponent software. J Pharm ICU. Chest 1995 Aug; 108 (2): 539-48
Biomed Anal 1991; 9 (6): 451-5 277. Newman LH, McDonald JC, Wallace PG, et al. Propofol
258. Blumer JL. Clinical pharmacology of midazolam in infants infusion for sedation in intensive care. Anaesthesia 1987
and children. Clin Pharmacokinet 1998 Jul; 35 (1): 37-47 Sep; 42 (9): 929-37
259. Reves JG, Fragen RJ, Vinik HR, et al. Midazolam: pharma- 278. McKeage K, Perry CM. Propofol: a review of its use in
cology and uses. Anesthesiology 1985 Mar; 62 (3): 310-24 intensive care sedation of adults. CNS Drugs 2003; 17 (4):
260. Allonen H, Ziegler G, Klotz U. Midazolam kinetics. Clin 235-72
Pharmacol Ther 1981; 30: 653-61 279. Simons PJ, Cockshott ID, Douglas EJ, et al. Disposition in
261. Greenblatt DJ, von Moltke LL, Ehrenberg BL, et al. male volunteers of a subanaesthetic intravenous dose of
Kinetics and dynamics of lorazepam during and after an oil in water emulsion of 14C-propofol. Xenobiotica
continuous intravenous infusion. Crit Care Med 2000 1988 Apr; 18 (4): 429-40
Aug; 28 (8): 2750-7 280. Veroli P, O’Kelly B, Bertrand F, et al. Extrahepatic me-
262. Harper KW, Collier PS, Dundee JW, et al. Age and nature tabolism of propofol in man during the anhepatic phase of
of operation influence the pharmacokinetics of mid- orthotopic liver transplantation. Br J Anaesth 1992 Feb;
azolam. Br J Anaesth 1985 Sep; 57 (9): 866-71 68 (2): 183-6
263. No authors listed. Clinical depression of the central ner- 281. Gray PA, Park GR, Cockshott ID, et al. Propofol metab-
vous system due to diazepam and chlordiazepoxide in olism in man during the anhepatic and reperfusion phases
relation to cigarette smoking and age. N Engl J Med 1973 of liver transplantation. Xenobiotica 1992 Jan; 22 (1):
Feb 8; 288 (6): 277-80 105-14
264. Olkkola KT, Ahonen J, Neuvonen PJ. The effects of the 282. Raoof AA, van Obbergh LJ, de Ville de GJ, et al. Extra-
systemic antimycotics, itraconazole and fluconazole, on hepatic glucuronidation of propofol in man: possible
the pharmacokinetics and pharmacodynamics of in- contribution of gut wall and kidney. Eur J Clin Pharmacol
travenous and oral midazolam. Anesth Analg 1996 Mar; 1996; 50 (1-2): 91-6
82 (3): 511-6 283. Crooks J, O’Malley K, Stevenson IH. Pharmacokinetics in
265. Greenblatt DJ, Shader RI, Abernethy DR. Drug therapy: the elderly. Clin Pharmacokin 1976; 1: 280-96
current status of benzodiazepines. N Engl J Med 1983 284. Bennett SN, McNeil MM, Bland LA, et al. Postoperative
Aug 11; 309 (6): 354-8 infections traced to contamination of an intravenous

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1914 Roberts et al.

anesthetic, propofol. N Engl J Med 1995 Jul 20; 333 (3): 303. Preckel B, Bolten J. Pharmacology of modern volatile an-
147-54 aesthetics. Best Pract Res Clin Anaesthesiol 2005 Sep; 19
285. Kress JP, Hall JB. Sedation in the mechanically ventilated (3): 331-48
patient. Crit Care Med 2006 Oct; 34 (10): 2541-6 304. Sakai EM, Connolly LA, Klauck JA. Inhalation anesthe-
286. Hall RI, Murphy JT, Moffitt EA, et al. A comparison of siology and volatile liquid anesthetics: focus on isoflurane,
the myocardial metabolic and haemodynamic changes desflurane, and sevoflurane. Pharmacotherapy 2005 Dec;
produced by propofol-sufentanil and enflurane-sufentanil 25 (12): 1773-88
anaesthesia for patients having coronary artery bypass 305. Ibrahim AE, Ghoneim MM, Kharasch ED, et al. Speed of
graft surgery. Can J Anaesth 1991 Nov; 38 (8): 996-1004 recovery and side-effect profile of sevoflurane sedation
287. Hall RI, Murphy JT, Landymore R, et al. Myocardial compared with midazolam. Anesthesiology 2001 Jan; 94
metabolic and hemodynamic changes during propofol (1): 87-94
anesthesia for cardiac surgery in patients with reduced 306. Berton J, Sargentini C, Nguyen JL, et al. AnaConDa re-
ventricular function. Anesth Analg 1993 Oct; 77 (4): 680-9 flection filter: bench and patient evaluation of safety and
volatile anesthetic conservation. Anesth Analg 2007 Jan;
288. Devlin JW, Lau AK, Tanios MA. Propofol-associated hy-
104 (1): 130-4
pertriglyceridemia and pancreatitis in the intensive care
unit: an analysis of frequency and risk factors. Pharma- 307. Maccioli GA, Cohen NH. General anesthesia in the in-
cotherapy 2005 Oct; 25 (10): 1348-52 tensive care unit? Is it ready for ‘‘prime time’’? Crit Care
Med 2005 Mar; 33 (3): 687-8
289. Roth MS, Martin AB, Katz JA. Nutritional implications of
prolonged propofol use. Am J Health Syst Pharm 1997 308. Kharasch ED. Adverse drug reactions with halogenated
Mar 15; 54 (6): 694-5 anesthetics. Clin Pharmacol Ther 2008 Jul; 84 (1): 158-62
290. Wong JM. Propofol infusion syndrome. Am J Ther 2010 309. Millane TA, Bennett ED, Grounds RM. Isoflurane and
Sep-Oct; 17 (5): 487-91 propofol for long-term sedation in the intensive care unit:
a crossover study. Anaesthesia 1992 Sep; 47 (9): 768-74
291. Orsini J, Nadkarni A, Chen J, et al. Propofol infusion
syndrome: case report and literature review. Am J Health 310. Hopkins PM. Malignant hyperthermia: pharmacology of
Syst Pharm 2009 May 15; 66 (10): 908-15 triggering. Br J Anaesth 2011 Jul; 107 (1): 48-56
311. Gelissen HP, Epema AH, Henning RH, et al. Inotropic
292. Diedrich DA, Brown DR. Analytic reviews: propofol in-
effects of propofol, thiopental, midazolam, etomidate,
fusion syndrome in the ICU. J Intensive Care Med 2011
and ketamine on isolated human atrial muscle. Anesthe-
Mar-Apr; 26 (2): 59-72
siology 1996 Feb; 84 (2): 397-403
293. Fodale V, La ME. Propofol infusion syndrome: an overview
312. Criado A, Maseda J, Navarro E, et al. Induction of an-
of a perplexing disease. Drug Saf 2008; 31 (4): 293-303
aesthesia with etomidate: haemodynamic study of 36
294. Kam PC, Cardone D. Propofol infusion syndrome. An- patients. Br J Anaesth 1980 Aug; 52 (8): 803-6
aesthesia 2007 Jul; 62 (7): 690-701
313. Ebert TJ, Muzi M, Berens R, et al. Sympathetic responses
295. Cremer OL, Moons KG, Bouman EA, et al. Long-term to induction of anesthesia in humans with propofol or
propofol infusion and cardiac failure in adult head- etomidate. Anesthesiology 1992; 76 (5): 725-33
injured patients. Lancet 2001 Jan 13; 357 (9250): 117-8 314. Dmello D, Taylor S, O’Brien J, et al. Outcomes of etomi-
296. Chukwuemeka A, Ko R, Ralph-Edwards A. Short-term date in severe sepsis and septic shock. Chest 2010 Dec; 138
low-dose propofol anaesthesia associated with severe (6): 1327-32
metabolic acidosis. Anaesth Intensive Care 2006 Oct; 34 315. Cotton BA, Guillamondegui OD, Fleming SB, et al. In-
(5): 651-5 creased risk of adrenal insufficiency following etomidate
297. Liolios A, Guerit JM, Scholtes JL, et al. Propofol infusion exposure in critically injured patients. Arch Surg 2008
syndrome associated with short-term large-dose infusion Jan; 143 (1): 62-7
during surgical anesthesia in an adult. Anesth Analg 2005 316. de Jong FH, Mallios C, Jansen C, et al. Etomidate sup-
Jun; 100 (6): 1804-6 presses adrenocortical function by inhibition of 11 beta-
298. Merz TM, Regli B, Rothen HU, et al. Propofol infusion hydroxylation. J Clin Endocrinol Metab 1984 Dec; 59 (6):
syndrome: a fatal case at a low infusion rate. Anesth An- 1143-7
alg 2006 Oct; 103 (4): 1050 317. Jackson Jr WL. Should we use etomidate as an induc-
299. Haase R, Sauer H, Eichler G. Lactic acidosis following tion agent for endotracheal intubation in patients with
short-term propofol infusion may be an early warning of septic shock? A critical appraisal. Chest 2005 Mar; 127
propofol infusion syndrome. J Neurosurg Anesthesiol (3): 1031-8
2005 Apr; 17 (2): 122-3 318. Meyer MR, Maurer HH. Absorption, distribution, me-
300. Soukup J, Scharff K, Kubosch K, et al. State of the art: tabolism and excretion pharmacogenomics of drugs of
sedation concepts with volatile anesthetics in critically ill abuse. Pharmacogenomics 2011 Feb; 12 (2): 215-33
patients. J Crit Care 2009 Dec; 24 (4): 535-44 319. Quibell R, Prommer EE, Mihalyo M, et al. Ketamine*.
301. Bracco D, Donatelli F. Volatile agents for ICU sedation? J Pain Symptom Manage 2011 Mar; 41 (3): 640-9
Intensive Care Med 2011 Jun; 37 (6): 895-7 320. Gunduz-Bruce H. The acute effects of NMDA antagonism:
302. Meiser A, Laubenthal H. Inhalational anaesthetics in the from the rodent to the human brain. Brain Res Rev 2009
ICU: theory and practice of inhalational sedation in the May; 60 (2): 279-86
ICU, economics, risk-benefit. Best Pract Res Clin An- 321. Chen RM, Chen TL, Lin YL, et al. Ketamine reduces nitric
aesthesiol 2005 Sep; 19 (3): 523-38 oxide biosynthesis in human umbilical vein endothelial

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
Sedation in the ICU 1915

cells by down-regulating endothelial nitric oxide synthase 339. Spies CD, Otter HE, Huske B, et al. Alcohol withdrawal
expression and intracellular calcium levels. Crit Care Med severity is decreased by symptom-orientated adjusted
2005 May; 33 (5): 1044-9 bolus therapy in the ICU. Intensive Care Med 2003 Dec;
322. Mazar J, Rogachev B, Shaked G, et al. Involvement of 29 (12): 2230-8
adenosine in the antiinflammatory action of ketamine. 340. Honey BL, Benefield RJ, Miller JL, et al. Alpha2-receptor
Anesthesiology 2005 Jun; 102 (6): 1174-81 agonists for treatment and prevention of iatrogenic opioid
323. Kawasaki T, Ogata M, Kawasaki C, et al. Ketamine sup- abstinence syndrome in critically ill patients. Ann
presses proinflammatory cytokine production in human Pharmacother 2009 Sep; 43 (9): 1506-11
whole blood in vitro. Anesth Analg 1999 Sep; 89 (3): 665-9 341. Hart GR, Anderson RJ. Withdrawal syndromes and the
324. Takenaka I, Ogata M, Koga K, et al. Ketamine suppresses cessation of antihypertensive therapy. Arch Intern Med
endotoxin-induced tumor necrosis factor alpha produc- 1981 Aug; 141 (9): 1125-7
tion in mice. Anesthesiology 1994 Feb; 80 (2): 402-8 342. Bohrer H, Bach A, Layer M, et al. Clonidine as a sedative
325. Nejati A, Moharari RS, Ashraf H, et al. Ketamine/ adjunct in intensive care. Intensive Care Med 1990; 16 (4):
propofol versus midazolam/fentanyl for procedural 265-6
sedation and analgesia in the emergency department: a 343. Gerlach AT, Dasta JF. Dexmedetomidine: an updated re-
randomized, prospective, double-blind trial. Acad Emerg view. Ann Pharmacother 2007 Feb; 41 (2): 245-52
Med 2011 Aug; 18 (8): 800-6 344. Hoy SM, Keating GM. Dexmedetomidine: a review of its
326. Hertle DN, Dreier JP, Woitzik J, et al. Effect of analgesics use for sedation in mechanically ventilated patients in an
and sedatives on the occurrence of spreading depolariza- intensive care setting and for procedural sedation. Drugs
tions accompanying acute brain injury. Brain 2012 Aug; 2011 Jul 30; 71 (11): 1481-501
135 (Pt 8): 2390-8 345. Bloor BC, Ward DS, Belleville JP, et al. Effects of in-
327. Taniguchi T, Yamamoto K. Anti-inflammatory effects of travenous dexmedetomidine in humans: II. Hemody-
intravenous anesthetics on endotoxemia. Mini Rev Med namic changes. Anesthesiology 1992 Dec; 77 (6): 1134-42
Chem 2005 Mar; 5 (3): 241-5 346. Dasta JF, Kane-Gill SL, Durtschi AJ. Comparing dexmede-
328. Jabre P, Combes X, Lapostolle F, et al. Etomidate versus tomidine prescribing patterns and safety in the naturalistic
ketamine for rapid sequence intubation in acutely ill setting versus published data. Ann Pharmacother 2004 Jul-
patients: a multicentre randomised controlled trial. Lan- Aug; 38 (7-8): 1130-5
cet 2009 Jul 25; 374 (9686): 293-300 347. Venn M, Newman J, Grounds M. A phase II study to
329. Kramer AH. Early ketamine to treat refractory status epi- evaluate the efficacy of dexmedetomidine for sedation in
lepticus. Neurocrit Care 2012 Apr; 16 (2): 299-305 the medical intensive care unit. Intensive Care Med 2003
330. Strayer RJ, Nelson LS. Adverse events associated with Feb; 29 (2): 201-7
ketamine for procedural sedation in adults. Am J Emerg 348. Jones GM, Murphy CV, Gerlach AT, et al. High-dose
Med 2008 Nov; 26 (9): 985-1028 dexmedetomidine for sedation in the intensive care unit:
331. Belopavlovic M, Buchthal A. Modification of ketamine- an evaluation of clinical efficacy and safety. Ann
induced intracranial hypertension in neurosurgical Pharmacother 2011 Jun; 45 (6): 740-7
patients by pretreatment with midazolam. Acta Anaes- 349. Dasta JF, McLaughlin TP, Mody SH, et al. Daily cost of
thesiol Scand 1982 Oct; 26 (5): 458-62 an intensive care unit day: the contribution of mechanical
332. Gardner AE, Dannemiller FJ, Dean D. Intracranial cere- ventilation. Crit Care Med 2005 Jun; 33 (6): 1266-71
brospinal fluid pressure in man during ketamine an- 350. Dasta JF, Kane-Gill S. Pharmacoeconomics of sedation in
esthesia. Anesth Analg 1972 Sep-Oct; 51 (5): 741-5 the ICU. Anesthesiol Clin 2011 Dec; 29 (4): 707-20, ix
333. Schwedler M, Miletich DJ, Albrecht RF. Cerebral blood 351. Anis AH, Wang XH, Leon H, et al. Economic evaluation
flow and metabolism following ketamine administration. of propofol for sedation of patients admitted to intensive
Can Anaesth Soc J 1982 May; 29 (3): 222-6 care units. Anesthesiology 2002 Jan; 96 (1): 196-201
334. Houston MC. Clonidine hydrochloride. South Med J 1982 352. Dasta JF, Kane-Gill SL, Pencina M, et al. A cost-
Jun; 75 (6): 713-9 minimization analysis of dexmedetomidine compared
335. Louis WJ, McNeil JJ, Anavekar SN, et al. Comparison of with midazolam for long-term sedation in the intensive
pharmacokinetics and pharmacodynamics of adrenoceptor care unit. Crit Care Med 2010 Feb; 38 (2): 497-503
agonists and antagonists as antihypertensive agents. J Car- 353. Al MJ, Hakkaart L, Tan SS, et al. Cost-consequence anal-
diovasc Pharmacol 1987; 10 Suppl. 12: S100-3 ysis of remifentanil-based analgo-sedation vs. conven-
336. Blaudszun G, Lysakowski C, Elia N, et al. Effect of peri- tional analgesia and sedation for patients on mechanical
operative systemic alpha2 agonists on postoperative ventilation in the Netherlands. Crit Care 2010; 14 (6):
morphine consumption and pain intensity: systematic re- R195
view and meta-analysis of randomized controlled trials. 354. Cox CE, Reed SD, Govert JA, et al. Economic evaluation
Anesthesiology 2012 Jun; 116 (6): 1312-22 of propofol and lorazepam for critically ill patients un-
337. Gowing L, Farrell M, Ali R, et al. Alpha2-adrenergic dergoing mechanical ventilation. Crit Care Med 2008
agonists for the management of opioid withdrawal. Co- Mar; 36 (3): 706-14
chrane Database Syst Rev 2009 Apr 15; (2): CD002024 355. Dasta JF, Jacobi J, Sesti AM, et al. Addition of dexmede-
338. Kraemer KL, Conigliaro J, Saitz R. Managing alcohol tomidine to standard sedation regimens after cardiac
withdrawal in the elderly. Drugs Aging 1999 Jun; 14 (6): surgery: an outcomes analysis. Pharmacotherapy 2006
409-25 Jun; 26 (6): 798-805

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)
1916 Roberts et al.

356. Muellejans B, Matthey T, Scholpp J, et al. Sedation in the 358. Barrientos-Vega R, Sanchez-Soria MM, Morales-Garcia C, et
intensive care unit with remifentanil/propofol versus al. Pharmacoeconomic assessment of propofol 2% used for
midazolam/fentanyl: a randomised, open-label, pharma- prolonged sedation. Crit Care Med 2001 Feb; 29 (2): 317-22
coeconomic trial. Crit Care 2006; 10 (3): R91
357. MacLaren R, Sullivan PW. Pharmacoeconomic modeling
of lorazepam, midazolam, and propofol for continuous Correspondence: Dr Richard I. Hall, Room 5452-Halifax
sedation in critically ill patients. Pharmacotherapy 2005 Infirmary, 1796 Summer St, Halifax, NS, B3H 3A7, Canada
Oct; 25 (10): 1319-28 E-mail: [email protected]

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (14)