CHEST Postgraduate Education Corner

CONTEMPORARY REVIEWS IN CRITICAL CARE MEDICINE

Patient-Focused Sedation
and Analgesia in the ICU*
Curtis N. Sessler, MD, FCCP; and Kimberly Varney, PharmD

Patient-focused sedation and analgesia in the ICU encompasses a strategy of comprehensive

structured management that matches initial evaluation, monitoring, medication selection, and
the use of protocols with patient characteristics and needs. This is best accomplished through
interdisciplinary management by physicians, nurses, and pharmacists. An early consideration is
that of the potential predisposing and precipitating factors, as well as prior sedative or analgesic
use, factors that may influence pharmacologic and supportive therapy. Frequent monitoring
with validated tools improves communication among clinicians and plays an important role in
detecting and treating pain and agitation while avoiding excessive or prolonged sedation.
Patient-focused management encompasses selecting medications best suited to patient charac-
teristics, including the presence of organ dysfunction that may influence drug metabolism or
excessive risk for side effects. The use of protocols to optimize drug therapy has emerged as a key
component of management, resulting in reductions in the duration of sedation, mechanical
ventilation, and ICU length of stay demonstrated with strategies to titrate medications to specific
targets, daily interruption of sedation, intermittent rather than continuous therapy, and analgesia-
based therapy. While much attention is paid to the initiation and maintenance of therapy, greater
emphasis must be placed on careful de-escalation of therapy in order to avoid analgesic or
sedative withdrawal. Finally, more work is needed to explore the relationship of critical illness
and sedation management with long-term psychological outcomes.
(CHEST 2008; 133:552–565)

Key words: analgesia; delirium; medications; protocols; sedation

Abbreviations: ATICE ⫽ Adaptation to the Intensive Care Environment; BIS ⫽ bispectral index; DIS ⫽ daily interruption
of sedation; EMG ⫽ electromyography; ICDSC ⫽ Intensive Care Delirium Screening Checklist; LOS ⫽ length of stay;
MSAT ⫽ Minnesota Sedation Assessment Tool; MV ⫽ mechanical ventilation; NMB ⫽ neuromuscular blockade;
PTSD ⫽ posttraumatic stress disorder; RASS ⫽ Richmond Agitation-Sedation Scale; RCT ⫽ randomized controlled trial;
SCCM ⫽ Society of Critical Care Medicine

C ritically ill patients, particularly those who are

receiving mechanical ventilation (MV), often have
ICU patients require IV sedative and analgesic med-
ications.5 It is important to recognize that these
pain, anxiety, dyspnea, and other forms of distress.1 underlying conditions, including delirium and delu-
Core principles of ICU care are to provide comfort, sional memories, as well as therapeutic interventions
to improve tolerance of the harsh ICU environment, may influence the likelihood of patients having long-
and to provide relief from distress. This is often term psychological effects.6
accomplished through identifying and correcting
predisposing and precipitating factors, applying non-
Interdisciplinary Management
pharmacologic measures to enhance comfort, and
administering sedative and analgesic medications, as It is noteworthy that effective sedation manage-
depicted in a conceptual framework in Figure 1.2 ment is best accomplished through intradisciplinary
Patients report that the greatest stressors they en- planning and practice. By combining nurses’ bedside
counter include pain, sleep deprivation, and the experience, skills, and continuous management of
presence of tubes in the nose and mouth, factors that sedation,7 pharmacists’ knowledge of medications,
are exceedingly common in the ICU setting.1,3,4 their interactions and proper role in critically ill
Accordingly, it is not surprising that the majority of patients,8 and physicians’ integration of sedation

552 Postgraduate Education Corner

 Figure 1. Conceptual framework for the interactions among underlying and causative factors, the
sensations of pain, anxiety, and delirium, pharmacologic and nonpharmacologic interventions, and the
state of a patient along the continuum from deep sedation to overt agitation. Reproduced with
permission from Sessler et al.2

issues into medical management,2,9 a comprehensive Initial Evaluation and Management

plan can be implemented on a consistent basis. An
overview of the principles of the management of An important starting place is to consider the
sedation and analgesia is displayed in Table 1. Effec- possible predisposing and precipitating factors for
tive management is patient focused, encompassing various forms of distress, plus issues that influence
medication selection to avoid adverse effects or management, as depicted in Figure 1. Underlying
prolonged effects from the parent drug or its active medical conditions such as chronic pain or arthritis,
metabolite(s), and titrating sedative and analgesic acute illness or injury, history of alcohol or substance
medications to specific targets for effectiveness (ie, abuse, and psychiatric illness can influence medication
tolerance of ICU environment, pain and anxiety selection. Other issues such as postoperative factors,
control, patient/ventilator synchrony) while avoiding ICU interventions such as MV, medications, and sim-
excessive or unnecessarily prolonged sedation that ple maneuvers like turning and suctioning, as well as
might lead to longer ICU length of stay and the sleep deprivation related to excessive noise and light,
accompanying complications of chronic critical illness. can all can play a role.2 Recognition and management
of these factors are crucial. Medication reconciliation
*From the Virginia Commonwealth University Health System, (ie, continuing a patient’s chronic home medication
Medical College of Virginia Hospitals, Richmond, VA. regimen) is frequently forgotten in hospitalized pa-
Dr Sessler is the Orhan Muren Professor of Medicine, Virginia
Commonwealth University Health System. tients. Antidepressants, anxiolytics, and antipsychotics
Dr. Sessler discloses receiving honoraria as a consultant from are medications of particular concern if not restarted.
Hospira, Inc. Severe withdrawal or rebound of chronic symptoms
Dr. Varney has no actual or potential conflicts of interest to disclose.
Manauscript received August 10, 2007; revision accepted Octo- can occur if these medications are withheld for too
ber 31, 2007. long, and can lead to serious comorbid issues. For
Reproduction of this article is prohibited without written permission example, patients who appear delirious or severely
from the American College of Chest Physicians (www.chestjournal.
org/misc/reprints.shtml). agitated in the ICU may become overmedicated with
Correspondence to: Curtis N. Sessler, MD, FCCP, Division of sedative drugs because their home medications have
Pulmonary and Critical Care Medicine, Box 980050, Virginia been temporarily discontinued. Recognition and man-
Commonwealth University Health System, Richmond, VA 23298;
e-mail: [email protected] agement of all of these factors early in their ICU stay
DOI: 10.1378/chest.07-2026 are crucial to optimal management of sedation.

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 553

 Table 1—Key Concepts for Management of Sedation dence that many cases of hypoactive or mixed delir-
and Analgesia ium in the ICU are related to the sedative effects of
1. Develop an interdisciplinary, structured approach for managing anxiolytic and analgesic drugs, particularly loraz-
sedation and analgesia in the ICU epam, that ICU caregivers administer15; thus, strat-
2. Perform patient assessment and optimize the ICU environment egies that emphasize using the lowest effective sed-
A. Identify predisposing and precipitating factors; manage ative drug dose may help avoid delirium.
treatable factors
Patient-focused sedation incorporates the concept
B. Identify outpatient medications (medication reconciliation),
particularly psychiatric and pain medications; restart that need for sedation and analgesia differs among
medications as appropriate patients and varies over time for individual patients.
C. Optimize patient comfort and tolerance of the ICU For example, patients who receive unusual forms of
environment ventilatory support such as high-frequency oscillatory
D. Optimize MV settings for patient/ventilator synchrony
ventilation or prone positioning may require deep
3. Regularly perform and document structured patient evaluation
and monitoring sedation, or even neuromuscular blockade (NMB), to
A. Establish and communicate treatment goals achieve ventilator synchrony.16 –18 Patients who are
B. Assess presence and severity of pain, as well as response to receiving NMB require adequate sedation and anal-
therapy gesia, and periodic cessation of the NMB agent to
C. Assess level of sedation using a validated sedation scale, as
assess adequacy of sedation and analgesia.19 In con-
well as response to therapy
D. Assess presence and severity of agitation using a validated trast, low tidal volume ventilation per se does not
agitation scale require increased sedative drug therapy in compari-
E. Identify delirium, and consider regular assessment of son to conventional ventilation for the ARDS.20 –22
delirium, using a validated delirium assessment instrument Converting the artificial airway from an endotracheal
4. Implement a structured patient-focused management strategy
tube to a tracheostomy tube has been associated with
A. Select analgesic and sedative drugs based upon patient
needs, drug allergies, organ dysfunction (particularly renal or use of less sedative and analgesic medication and
hepatic dysfunction), need for rapid onset and/or offset of correspondingly fewer hours of deep sedation.23 Alco-
action, anticipated duration of therapy, and prior response to hol and substance abuse is associated with greater
therapy sedative drug requirements.24 Importantly, require-
B. Focus first on analgesia, then sedation
ments for sedation are dynamic, generally declining
C. Titrate analgesic and sedative drugs to a defined target,
using the lowest effective dose as illness improves, and thus must be reassessed
D. Implement a structured strategy to avoid accumulation of frequently.
medications/metabolites: utilize scheduled interruption, or
intermittent dosing of analgesic and sedative drugs
E. Evaluate and manage severe agitation, including search for Evaluation of Pain, Sedation, Agitation,
causative factors, and perform rapid tranquilization
and Delirium
F. Identify delirium, correct precipitating factors, and treat with
haloperidol or other neuroleptic drugs
Bedside evaluation of the level of consciousness or
G. Avoid potential adverse effects of analgesic and sedative
drugs; quickly identify and manage adverse effects that arousability, cognitive function, presence and inten-
occur sity of pain, and presence and intensity of agitation
5. Recognize and take steps to ameliorate analgesic and sedative is an integral component of daily patient care.2,9,25
drug withdrawal during de-escalation of therapy The detection and quantification of pain or agitation
is useful to prompt the initiation or escalation of
therapy as well as the reevaluation for reversible
causes.26 The routine determination of the level of
Delirium, an acute reversible disorder of attention
consciousness, which often incorporates the domains
and cognition, is a common occurrence among crit-
of arousal and cognition, helps guide sedative drug
ically ill patients, and may be a marker for worse
dosing and avoidance of excessively deep sedation.
outcomes, including increased incidence of posttrau-
Thus, utilization of tools—sedation scales and pain
matic stress disorder (PTSD) and increased long-
evaluation instruments—is important for patient-
term mortality.10,11 While classical hyperactive de-
focused therapy, typically by establishing a target
lirium is easily recognized by manifestations of
level of sedation to which medications are titrated to
intermittent agitation, hallucinations, and disrup-
achieve, and to minimize pain.
tive behavior, a hypoactive form appears to be more
common in the ICU setting.12 A wide variety of
Pain Evaluation
medical conditions (including CNS disorders, infec-
tions, hypoxia, pain, withdrawal syndromes, electro- Evaluation of pain is relatively straightforward for
lyte and metabolic disorders, and various organ patients who are alert enough to “self-report” by
dysfunctions) as well as a long list of medications can speaking, nodding, or pointing in response to ques-
precipitate delirium.2,11,13,14 There is emerging evi- tions about severity of pain. This is often performed

554 Postgraduate Education Corner

using an instrument such as a 10-cm visual analog with the Ramsay Sedation Scale, RASS, ATICE, and
scale, or a scale with extremes of measurement an- MSAT. In some scales, cognition—as the ability to
chored by numerical (0 to 10), descriptive (“no pain” to follow a command (ie, RASS, ATICE)—and sustain-
“worst pain ever”), or diagrammatic (smiling face to ability (ie, RASS) are included in the testing.41,42
crying face) variables, the patient indicating their Additionally, agitation is identified and graded in
level of pain.27 Asking about pain should be per- severity in Sedation Agitation Scale, Motor Activity
formed frequently, particularly prior to administer- Assessment Scale, RASS, and ATICE.39 – 42 Finally,
ing sedatives or when these agents are stopped widespread integration into ICU practice has been
temporarily. Assessment of pain is less reliable and documented for RASS, MSAT, and ATICE.41,44 – 47
less valid when inferred through observation of Implementation of sedation evaluation with a scale
patient behaviors, although recent instruments such is an integral component of many treatment algo-
as the Critical Care Pain Observation Tool28 show rithms, and has been documented to result in
promise. The components of Critical Care Pain Obser- more precise dosing, reduced sedative and analgesic
vation Tool and other similar tools29 –31 are based on drug use, shorter duration of MV, and reduced need
recognition of common behaviors, such as facial for vasopressor therapy,48 as well as reduced inci-
grimacing, restless body movement, rigid limbs, and dence of oversedation.49
patient/ventilator asynchrony observed during pain-
ful procedures and validated against self-report.32,33 Objective Measurement of Brain Activity
While hypertension, tachycardia, and tachypnea are Objective assessment of brain activity can be
also commonly observed during painful stimuli, their performed using EEG signals processed by propri-
lack of specificity for pain34 has dampened enthusi- etary algorithms such as with the bispectral index
asm for routine use. These symptoms may be more (BIS),50 patient state index,51 cerebral state index,52
telling if combined with a positive history of chronic and Narcotrend index53 systems to yield a single
or acute pain, or use of chronic pain medications. numerical value from 0 (complete EEG suppression)
to 100 (awake). These monitors offer advantages of
Sedation Evaluation using objective physiologic parameters, a simplified
numerical display, and near-continuous measure-
The level of sedation is assessed in many ICUs
ment, yet are rarely utilized in the clinical ICU
using a sedation scale; however, surveys indicate they
setting today.50 Factors such as electromyography
are used by fewer than one half of practitioners,35
(EMG) or electrical current-related artifact, which
ICUs,36 and consecutive patients receiving MV.37
are less problematic in the operating room setting,
Useful features of a sedation scale include rigorous
increase variability of BIS in the critically ill patient.
multidisciplinary development; ease of administra-
In fact, the administration of a NMB agent to
tion, recall, and interpretation; well-defined discrete
sedated ICU patients led to remarkable decreases in
criteria for each level; sufficiency sedation levels for
BIS, averaging 24 U and 35 U in two studies,54,55
effective drug titration; assessment of agitation; dem-
demonstrating the potential importance of EMG
onstration of interrater reliability for relevant patient
interference. While EMG artifact may account in
populations; and evidence of validity.25 Although a
part for the wide variability in correlation (r2 ⫽ 0.21
number of sedation scales have been developed for
to 0.93) between BIS and level of sedation as judged
ICU use, the following sedation scales satisfy many
by various sedation scales,50 the use of newer algo-
of these characteristics and have been tested for
rithms designed to reduce EMG influence may not
interrater reliability in multiple patient populations at
eliminate BIS variability.56 Experts recommend us-
multiple hospitals and validated against numerous mea-
ing BIS for selected cases, such as monitoring the
sures: the Ramsay Sedation Scale,38 the Sedation Agi-
level of consciousness of patients treated with NMB
tation Scale,39 the Motor Activity Assessment Scale,40
agents and for deep (pentobarbital-induced) coma57;
the Richmond Agitation-Sedation Scale (RASS),41
however, routine use must await studies that dem-
the Adaptation to the Intensive Care Environment
onstrate added value from better outcomes and/or
(ATICE) instrument,42 and the Minnesota Sedation
cost savings. Additional approaches to brain moni-
Assessment Tool (MSAT).43 A common approach to
toring in the ICU include response entropy and state
testing level of consciousness is to observe the
entropy,58 and auditory-evoked potentials.59,60
patient to see if they are awake, and if not, then to
stimulate the patient using auditory stimulation
Delirium Evaluation
(speaking to the patient), followed by more intensive
(physical) stimulation if there is no response to Delirium is associated with worse outcomes10,11
auditory stimuli. The patient’s response to increasing and has been frequently found in ICU patients when
levels of stimulation is noted and a score assigned, as using some10,61,62 but not all assessment tools11,63– 65

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 555

and depending on the patient population. Detection tract or following subcutaneous or IM injection is
of delirium in the critically ill patient has been less reliable, and precise titration to the target effect
challenging because inability to speak and often to is more accurate. However, administration of trans-
write impairs testing, yet tools for use in ICU dermal or enteral medications may have a role in
patients have been developed.66 Many validated some patients, particularly those who require longer-
tools are interventional, such as the Confusion As- term sedation and analgesia management, resulting
sessment Method for the ICU,61 or the Cognitive in elimination or reduction in IV dosages.82 Mea-
Test for Delirium67 (or its abbreviated version),68 in sures must be taken to avoid long-term sedative or
which a rater directly tests state of consciousness, opioid medication dependence after ICU discharge;
comprehension, memory, attention, and vigilance. explicit plans for cessation or scheduled tapering of
Other validated instruments use observation with a these drugs are necessary as the patient leaves the
checklist of behaviors such as level of consciousness, ICU, particularly for patients with a history of alco-
inattention, disorientation, hallucinations, agitation, hol or substance dependency.
inappropriate mood, sleep-wake cycle disturbance,
and symptom fluctuation as in the Intensive Care
Delirium Screening Checklist (ICDSC)63 or the Medications for Sedation and Analgesia
Nursing Rating Scale for ICU Delirium.67 The Con-
The two major classes of medications used to
fusion Assessment Method for the ICU and ICDSC
promote comfort and tolerance of the ICU environ-
are more recently developed and have been exten-
ment are the sedative-hypnotic agents and opioid
sively validated. The ICDSC also allows identifica-
analgesics, which provide anxiolysis, sedation, amne-
tion of “subsyndromal” delirium that may progress to
sia, and analgesia to the patient. General consider-
overt delirium.69 Although routine screening for the
ations for medication selection include effectiveness;
presence of delirium is recommended in the 2002
factors related to duration, onset, and offset of effect;
Society of Critical Care Medicine (SCCM) guide-
presence of active metabolites; adverse effects; costs
lines for sedative and analgesic therapy,9 surveys
related to drug acquisition as well as to duration of
indicate that delirium testing of ICU patients with a
sedation; and any recent history of opioid or benzo-
validated tool is rarely performed.35,70
diazepine use. Characteristics of commonly used IV
Other components of sedation-related monitoring of
administered sedative and analgesic drugs are dis-
the ICU patient include detection and quantification of
played in Table 2.
conditions that influence sedation management deci-
sions, including agitation,39 – 42 patient movement as
Sedative Medications
a surrogate for agitation,43,71,72 and patient/ventilator
asynchrony.73,74 Future approaches will likely inte- The benzodiazepines, midazolam and lorazepam, are
grate multiple measurements in order to optimize commonly administered by continuous infusion or in-
patient comfort and tolerance of the ICU setting. termittent injection in the North American ICUs,35,83
whereas lorazepam is rarely used and midazolam
enjoys widespread use in Europe.37,84 Lorazepam
Managing Sedation and Analgesia given by intermittent boluses or continuous IV infu-
sion was recommended in the 2002 SCCM consen-
Pharmacologic therapy is required in the majority sus guidelines as the preferred sedative drug for ICU
of ICU patients, particularly those who are receiving patients who require prolonged therapy, whereas
MV.5 Medications are often administered by contin- midazolam was recommended only for short-term
uous infusion for ease of use and to provide a smooth (⬍ 48 h) sedation because of concerns for unpredict-
course; however, continuous infusion has been linked able awakening observed after prolonged infusion.9
to prolonged sedation and longer length of ICU Some of the delayed emergence with midazolam
stay.75 Accordingly, intermittent therapy or provision may be attributable to accumulation of the parent
of schedule daily interruption of sedation (DIS) is compound in hepatic failure, or an active metabo-
often employed to avoid excessive and prolonged lite, ␣-hydroxymidazolam, which is cleared by the
effects.76 –78 Often, multiple medications are admin- kidney and may lead to prolonged sedation in pa-
istered in order to treat both pain and anxiety, or to tients with renal insufficiency.85,86 In a series of
provide synergy, thus allowing reduced dosing of consecutive patients with prolonged sedation after
medications.5,79,80 Several studies suggest that focus- cessation of midazolam infusion, many patients had
ing first on providing analgesia rather than initially elevated serum levels of ␣-hydroxymidazolam (all
on anxiolysis may provide more effective79 and patients had serum creatinine ⬎ 1.5 mg/dL), or
shorter duration of MV.81 In most cases, IV admin- detectable levels of midazolam many hours after
istration is desirable since absorption from the GI infusion discontinuation.86 There are relative few

556 Postgraduate Education Corner

 Table 2—IV Sedative and Opioid Analgesic Medications Used in Adult ICU Patients*

Relative Daily
Drug/Class Elimination Onset/Duration Dosing (IV) Concentration Advantages Disadvantages Cost†

www.chestjournal.org
Lorazepam/ Hepatic conjugation to
5–20 min/6–8 h; up LD: 2–4 mg IV push MD: 2–6 100 mg/100 mL Inexpensive, longer Propylene glycol toxicity at high $$
benzodiazepine inactive metabolite to 24–72 h in mg IV q4h-q6h; infusion: D5W only half-life doses (anion gap metabolic
elderly/cirrhosis/ 1–10 mg/h; start low in acidosis, renal insufficiency)
ESRD elderly
Midazolam/ Cytochrome P450 3A4; 5–10 min/1–4 h LD: 2–5 mg IV push MD; 1–20 100 mg/100 mL Shorter acting if Many drug interactions, may $$$
benzodiazepine active metabolite (longer in ESRD/ mg/h; start low in elderly NS or D5W preserved organ increase midazolam levels,
excreted renally CHF/liver failure) funct ion; fast active metabolite accumulates
onset in renal failure
Propofol Conjugation 30–50 s/ MD: 5–150 ␮g/kg/min Premixed (10 Short acting 2BP, increase serum $$$$
approximately mg/mL) triglyceride, pancreatitis,
3–10 min (dose propofol infusion syndrome,
dependent) zinc depletion
Dexmedetomidine Hepatic Cytochrome Immediate/ LD: 0.5–1 ␮g/kg over 10 min; 100 ␮g/50 mL Very short duration; 2BP, 2HR; not approved for $$$$$
P450 and approximately 6 MD: 0.2–0.7 ␮g/kg/h for 24 h NS only has some analgesic use ⬎ 24 h, some studies use
glucuronidation min (longer in properties longer
liver failure)
Morphine Conjugation; active 5–10 min/2–4 h LD: 2–4 mg IV push MD: 2–30 100 mg/100 mL Reduces tachypnea 2BP, respiratory depression, $
sulfate/opioid metabolite excreted mg/h for ventilated patients NS or D5W accumulation in hepatic/renal
analgesic renally failure
Fentanyl/opioid Cytochrome P450 3A4 1–2 min/2–4 h LD: 25–50 ␮g IV push MD: 1.25 or 2.5 mg/ Less hypotension 3A4 inhibitors may increase $$
analgesic (longer in liver 0.7–10 ␮g/kg/h for ventilated 250 mL NS or than morphine fentanyl; fever will increase
failure) patients D5W patch fentanyl levels by 30%
Hydromorphone/opioid Hepatic 5–10 min/2–4 h LD: 0.2–0.6 mg IV push MD: 100 mg/100 mL May work if patients Respiratory depression, caution $
analgesic 0.5–3 mg/h NS or D5W are tolerant to in nonventilated patient;
morphine/fentanyl highly addictive
Alfentanil/opioid Hepatic; active 1 min/30–60 min LD: 50–75 ␮g/kg slowly over 10 mg/250 mL Very short-acting 2HR, 2BP, 1ICP; 3A4 $$
analgesic metabolites excreted (dose dependent) 3–5 min; MD: 0.5–3 ␮g/kg/ NS or D5W agent inhibitors may increase levels
renally min (usual 1–1.5 ␮g/kg/min) of alfentanil
Remifentanil/opioid Tissue esterases 1–3 min/10–20 min LD: 1 ␮g/kg over 1 min MD: 5 mg/250 mL No accumulation in 2HR, 2BP, 1ICP $$$
analgesic 0.6–15 ␮g/kg/h for MV NS or D5W hepatic or renal
(unlabeled use); use ideal failure
body weight if ⬎ 30% over
ideal body weight
Sufentanil/opioid Hepatic 1–3 min/dose- LD: 1–2 ␮g/kg slowly over 3–5 250 ␮g/250 mL 2HR, 2BP, 1ICP $
analgesic dependent min; MD: 8–50 ␮g as needed D5W; variable
duration stability in NS
*ESRD ⫽ end-stage renal disease; LD ⫽ loading dose; MD ⫽ maintenance dose; D5W ⫽ 5% dextrose in water; CHF ⫽ congestive heart failure; NS ⫽ normal saline solution; 2BP ⫽ hypotension;
2HR ⫽ bradycardia; 1ICP ⫽ increased intracranial pressure.

CHEST / 133 / 2 / FEBRUARY, 2008

†Dollar signs represent orbitrary scale of medication costs, ranging from $ ⫽ very low cost, to $$$$$ ⫽ very high cost.

557
randomized controlled trials (RCTs) in which loraz- gated. In small studies of short-term use in the
epam and midazolam are compared for long-term postoperative or trauma patients, dexmedetomidine
(⬎ 72 h) infusion. No significant difference in time has been shown to decrease opiate use, and to
to awakening was noted in two studies,87,88 whereas facilitate extubation in patients who have failed
Barr and coworkers89 found significantly shorter previous ventilator weaning attempts due to severe
emergence time and time to extubation for midazo- agitation.96 Case reports have also documented suc-
lam vs lorazepam infusion in a double-blinded RCT cess in preventing alcohol withdrawal in patients in
of postoperative patients without significant renal, the perioperative period.97 Drug-related adverse ef-
hepatic, or neurologic impairment. Additionally, pro- fects are primarily cardiovascular including hypoten-
longed high-dose administration of lorazepam can sion and bradycardia,98 particularly when loading
result in accumulation of the vehicle, propylene doses are used. Tolerance to the drug has been seen
glycol, resulting in worsening renal function, meta-
and there are concerns for a rebound effect when
bolic acidosis, and altered mental status.90 Toxicity is
used beyond 24 to 48 h. Most studies find that
typically observed after prolonged (⬎ 7 d), high-dose
higher-than-recommended doses are needed for ef-
(average of 14 mg/h), continuous lorazepam infusion,
and can be recognized by measuring an increased ficacy but that a ceiling dose effect is reported at a
osmolal gap.90 Lorazepam infusion was identified as dose approximately 1.5 ␮g/kg/h.99 Dystonia has
an independent risk factor for transition to delirium been reported and may be due to its effect on
among ICU patients receiving MV15; however, ex- acetylcholine release.99 Acquisition cost is higher
amination of the odds ratios for other sedative and than for other sedative agents, although one ret-
opioid analgesics agents used less frequently in the rospective outcomes study100 of ⬎ 10,000 cardiac
study suggests it may occur with other agents. In surgery patients suggests favorable clinical and
light of drug accumulation with continuous infusion, economic outcomes when dexmedetomidine is
and concerns for propylene glycol toxicity, long-term added to midazolam and propofol for sedation.
high-dose lorazepam should be used with caution. Publication of additional research will help to
Midazolam may be a better alternative if hepatic and clarify new roles for this drug.
renal function is normal.
Despite consensus recommendations that admin- Opioid Analgesic Medications
istration of propofol be limited to ⬍ 24 to 48 h,9 it
remains widely used worldwide.5,35,83,84 Propofol in- A variety of opioids used by IV administration in
fusion has been compared to midazolam infusion for adults are available for use in ICUs throughout the
sedation in the ICU in many RCTs. Walder and world; many of these are compared in Table 2.
colleagues91 demonstrated more effective sedation Surveys and prospective surveillance studies35–37,83,84
(higher percentage of time) and shorter weaning indicate widespread use of fentanyl and morphine
time for short-term (⬍ 36 h) infusion but higher sulfate, although sufentanil enjoys considerable use
adverse event rates with propofol compared to mi- in Europe.37,84 The 2002 SCCM guidelines recom-
dazolam in a systematic review of 27 RCTs. Duration mend the use of fentanyl, hydromorphone, or mor-
of MV was similar for propofol and midazolam phine if an IV opioid analgesic is required.9 Opioids
infusions, however, when DIS was routinely used function through stimulation of receptors, principally
with both drugs.77 In contrast, duration of MV was via the ␮1 and ␮2 opioid receptors. All opioids
significantly shorter with propofol infusion and DIS produce a dose-dependent respiratory depression;
compared to intermittent lorazepam in a recent other common side effects include muscle rigidity,
RCT.78 Propofol infusion has been linked to a num- hypotension, delayed GI transit, nausea, pruritus,
ber of adverse effects,92 including hypertriglyceride- and urinary retention.27 There are important differ-
mia,93 dose-dependent hypotension,94 and the ence in regards to lipid solubility, volume of distri-
propofol infusion syndrome, a rare clinical syndrome bution, and metabolism that are reflected in data
of rhabdomyolysis, metabolic acidosis, renal failure, displayed in Table 2. Particularly noteworthy is that
and cardiac failure after high doses of propofol.95 morphine is metabolized to several active metabo-
Dexmedetomidine is a selective ␣2-adrenergic re- lites, including morphine-6-glucuronide that is more
ceptor agonist with a short half-life of approximately active than the parent compound, and accumulates
2.3 h, and has sedative, analgesic, anxiolytic, and in renal failure potentially resulting in prolonged
sympatholytic effects without depressing respiratory sedation and respiratory depression.101 Morphine
drive. Dexmedetomidine currently has limited ap- should be avoided in patients who have renal insuf-
proval by the US Food and Drug Administration for ficiency. Fentanyl has a rapid onset of action as a
sedation in postoperative patients of ⬍ 24 h dura- result of high lipophilicity and a short duration of
tion, although other ICU settings are being investi- action from redistribution.27 Elimination is delayed,

558 Postgraduate Education Corner

however, with prolonged administration as a result of tracheostomy, shorten the ICU and hospital length
its large volume of distribution. of stay, decrease the likelihood of having ICU-related
Often analgesic medications are added to a seda- medical complications, and reduce costs of hospitaliza-
tive infusion, so-called “co-sedation” or “analgoseda- tion. A variety of strategies have been employed,
tion,” with reduced doses of both agents and in some including using medications with a shorter half-life,
cases more effective therapy.79 European studies titrating medications to end-points identified by
demonstrate shorter duration of MV with an analgesic- sedation scales, mandating temporary cessation of
based vs sedative-based protocol.81 Remifentanil, a sedative drug infusion, use of intermittent sedative
drug that has organ-independent metabolism, was therapy, and employing algorithms that proactively
successfully used in this trial81; however, other clin- identify pain, agitation, and/or patient/ventilator
ical trials demonstrate similar results between remifen- asynchrony. Several studies that tested strategies
tanil and fentanyl, except for a greater incidence of pain using a prospective controlled study design deserve
during drug de-escalation with remifentanil.102 further comment (Table 3).
Following demonstration that continuous infu-
Antipsychotic Agents sion of sedatives and analgesics was associated
Therapy with antipsychotic agents such as haloper- with prolonged ventilator time and with longer
idol may be necessary for effective management of ICU and hospital LOS,75 Brook and colleagues76
delirium and agitation. For treatment of acute agita- conducted a single-center RCT comparing a nursing-
tion, the butyrophenone haloperidol can be admin- implemented algorithm with conventional practice.
istered in escalating doses from 1 mg to as much as This algorithm emphasized the following: (1) early
20 mg by IV injection, often in combinations with a detection of pain, (2) use of intermittent therapy
benzodiazepine, until calmness is achieved.103 Halo- with fentanyl or lorazepam, reserving continuous
peridol has been associated with extrapyramidal ef- infusions for patients who had inadequate response
fects, and cardiac conduction abnormalities, specifi- to intermittent therapy, and (3) de-escalation of
cally prolonged QTc-associated torsades de pointes continuous infusions to intermittent therapy. They
arrhythmia.104 These tend to be dose-related adverse demonstrated shorter duration of MV, shorter ICU
effects. A baseline ECG should be documented and and hospital LOS, and lower rates of tracheostomy.
periodically repeated with continued use of the drug.
In addition to maintaining normal magnesium, po- Daily Interruption of Sedation
tassium, and calcium levels, it is important to consult
Kress and colleagues77 tested a strategy of DIS
with a pharmacist regarding drug interactions that
compared to usual practice in an RCT, also compar-
could potentiate cardiac arrhythmias (ie, amiodarone
ing midazolam to propofol infusions in a 2 ⫻ 2
and other antiarrhythmics, antifungal azoles, quino-
factorial design. All DIS patients had sedative and
lones, macrolides, etc.). Haloperidol may be admin-
analgesic (morphine sulfate) infusions discontinued
istered as a scheduled medication for management
each morning until the patient was able to follow
of delirium. Interestingly, in a retrospective study,
three or more of four simple commands or became
use of haloperidol in ICU patients receiving MV was
agitated. Infusions were restarted at one half the
associated with a lower mortality rate in a dose-
original rate. They demonstrated significant reduc-
response fashion than nonuse, although the explana-
tions in duration of MV and shorter ICU LOS, as
tion(s) for this is not clear.105 While there is interest
well as fewer diagnostic studies for unexplained
in examining newer antipsychotic agents for manage-
altered mental status. There was no difference in any
ment of delirium, only olanzapine has been directly
outcomes between the midazolam and propofol
compared to haloperidol thus far. Olanzapine was
groups, although DIS resulted in significantly lower
found to be of similar efficacy to haloperidol in
daily doses of midazolam and morphine but not
controlling delirium, but with fewer extrapyramidal
propofol. This observation implies that more rapid
side effects in a small prospective trial.106
extubation and ICU discharge may be related to a
combination of reduced drug accumulation, and
Protocols and Algorithms To Improve additional opportunities for initiating weaning from
Sedation Management MV.110 This strategy has also been linked to a
reduction in ICU complications, largely as a result of
Structured approaches to the management of se- a shorter ICU LOS.111
dation and analgesia have been demonstrated in In a two-center RCT, Carson et al78 compared
prospective studies26,47,76 –79,81,107–109 to reduce side propofol infusion vs intermittent therapy with loraz-
effects of medications, decrease unnecessary testing, epam; both groups received DIS and morphine
reduce the duration of MV and the frequency of sulfate for analgesia. The continuous infusion propo-

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 559

560
Table 3—Prospective Trials of Sedation Protocols in Adult ICU Patients*

Key Components of
Source Patients and Setting Study Design Protocol Major Findings Comments
76
Brook et al 321 MV medical ICU RCT: protocol-directed Nursing-directed protocol Protocol group had shorter duration of MV (56 h vs 117 h,
patients, US sedation vs non- that emphasizes using p ⫽ 0.008); shorter ICU LOS (5.7 d vs 7.5d, p ⫽ 0.013);
university hospital protocol–directed intermittent therapy; shorter hospital LOS (14.0 d vs 19.9 d, p ⫽ 0.003);
sedation (control) fentanyl, lorazepam lower tracheostomy rate (6.2% vs 13.2%, p ⫽ 0.038)
Kress et al77 128 MV medical ICU RCT: intervention vs Intervention group Intervention group had shorter duration of MV (4.9 d vs 2 ⫻ 2 factorial design;
patients; US control received DIS 7.3 d, p ⫽ 0.004); shorter ICU LOS (6.4 d vs 9.9 d, patients also
university hospital p ⫽ 0.02); and fewer diagnostic tests for changes in randomized to
mental status (9% vs 27%, p ⫽ 0.02) receive midazolam
or propofol
MacLaren et al107 158 MV patients in Prospective two-phase Evidence-based sedation Protocol group had less “discomfort” (11% vs 22 0.4%, Protocol adherence ⫽
medical-surgical- study: empiric before/ and analgesia protocol p ⬍ 0.001); less pain (5.9% vs 9.6%, p ⬍ 0.05); lower 83.7%
neurologic protocol after hourly sedation cost (Canadian $5.68 vs Canadian $7.69,
Canadian ICU p ⬍ 0.01); trend for longer sedation duration (122.7 h vs
88.0 h, p ⬍ 0.1); trend for longer duration of MV (61.6
h vs 39.1 h, p ⫽ 0.13)
Mascia et al108 158 MV medical and Prospective two-phase Guidelines based upon Guideline group had lower direct drug costs; shorter
surgical ICU study: baseline before/ “rational cost-effective duration of MV (167 h vs 317 h)†; shorter ICU
patients, US tertiary guidelines after use of drugs” LOS (9.2.d vs 19.1d)†; shorter hospital LOS (19.1 d vs
care university 34.3 d)†
hospital
Brattebo et al109 285 MV surgical ICU Prospective two-phase Small-scale rapid-cycle Protocol group had shorter duration of MV (5.3 d vs 7.4
patients, Norwegian study: baseline before/ improvement model; d)†; trend for shorter ICU LOS (8.3 d vs 9.3 d)†
university hospital guideline after protocol, sedation scale
De Jonghe et al47 102 MV medical ICU Prospective two-phase Algorithm: regular Algorithm group had shorter time to arousal (2 d vs 4 d,
patients, French study: control before/ assessment of p ⫽ 0.006); shorter duration of MV (4.4 d vs 10.3 d,
university-affiliated algorithm after consciousness and p ⫽ 0.014)
hospital tolerance to ICU
environment with goal
of tolerance and high
LOC
Breen et al81 105 MV medical- 15-center RCT: analgesic Titration of analgesic Analgesic-based group had shorter time from weaning to
surgical ICU (remifentanil based) or (remifentanil, fentanyl, extubation (0.9 h vs 27.5 h, p ⬍ 0.001); shorter duration
patients, 10 sedative (midazolam or morphine) and of MV (94 h vs 147.5 h, p ⫽ 0.033)
European countries based) sedative (midazolam)
drugs to targets
Carson et al78 132 MV, sedated RCT: lorazepam by Daily interruption of Continuous infusion propofol group had shorter duration
medical ICU intermittent bolus, or sedation was of MV (5.8 d vs 8.4 d, p ⫽ 0.04); trend for more
patients, two US propofol by continuous performed in both ventilator-free survival (18.5 d vs 10.2 d, p ⫽ 0.06)
university hospitals infusion groups
(Continued)

Postgraduate Education Corner

 fol with DIS group had significantly shorter duration

duration of MV not
of MV. Thus, DIS is emerging as a useful technique

Small sample size;

trend for more
Comments
to shorten duration of MV, and this “sedation vaca-

co-sedation;
ileus with
tion” has been widely embraced.112

reported
Questions arose regarding the neuropsychiatric
impact of repeated abrupt awakening on critically ill
patients,113 yet subsequent research has demon-
strated that patients randomized to DIS have fewer
symptoms of PTSD compared to control subjects.114
Protocol group had lower incidence of pain (42% vs 63%,

patient/ventilator asynchrony (0.4/h vs 1.0/h, p ⬍ 0.05)

sedation (4.2 h vs 9.1 h, p ⬍ 0.002); fewer episodes of
p ⫽ 0.002); lower incidence of agitation (12% vs 29%,

p ⬍ 0.05); fewer nosocomial infections (8% vs 17%,

Interruption of sedation is associated with a surge in
p ⫽ 0.002); shorter duration of MV (65 h vs 120 h,

Co-sedation group had fewer hour with “off- target” circulating catecholamines; however, patients with
coronary artery disease subjected to DIS had no
increase in cardiac ischemia when taken off seda-
tion.115 These studies are encouraging for the safety
Major Findings

of widespread use of DIS, although caution for selected

patients, such as those with hypertensive crisis or status
asthmaticus, who might have significant consequences
from the resulting stress response or patient/ventila-
tor asynchrony, should be exercised.110 Finally, there
are theoretical concerns with regards to the use of
DIS in patients who have a history of alcohol abuse
p ⬍ 0.05)

because precipitation of withdrawal could occur if

the patient is not otherwise receiving medications
that protect for alcohol withdrawal.
Other protocol-based approaches to sedation man-
Table 3—Continued

agement have been demonstrated in a sequential

study design with comparison to an earlier control
systematic evaluation

protocol; midazolam,
of pain and agitation
Key Components of

Protocol emphasis on

period, to reduce duration of MV,47,81,108,109 improve

Nurse-implemented
Protocol

the quality of sedation,107 reduce patient/ventilator

asynchrony,79 and reduce drug costs,79 although one
fentanyl

protocol that increased use of lorazepam was associ-

ated with a trend for longer ventilator duration.107
Common themes for effective protocols include
targeting pain, agitation, and intolerance of the ICU
environment,26,47 and focusing more on analgesic
benzodiazepine, or co-
study: control before/

therapy than sedation.79,81

benzodiazepine plus
Prospective two phase
Study Design

RCT: sedation with

protocol after

sedation with

Sedative and Analgesic Drug Withdrawal

opiate

Development of signs and symptoms of acute

opioid and sedative drug withdrawal syndrome can
follow long-term administration of these medications
Patients and Setting

in ICU patients. This phenomenon has been partic-

30 MV medical ICU
university hospital

university hospital
patients, French

ularly well studied in critically ill children.116 –118

230 MV medical-
surgical ICU

patients, US

Cammarano et al119 reported that one third of adult

trauma patients who received ⬎ 1 week of ICU hospi-
†Indicates p values not reported.
*LOC ⫽ level of consciousness.

talization had clear evidence of acute withdrawal once

drugs were discontinued. Longer duration and higher
doses of opioids, benzodiazepines, and propofol were
associated with withdrawal, as was less haloperidol
26

administration. Recently, withdrawal was detected in

Richman et al79
Chanques et al

13.2% of ICU patients and linked to delirium in the

Source

ICU.6 Research suggests that CNS and sympathetic

nervous system markers for withdrawal can peak
with 6 h of cessation of sedative (midazolam, propo-

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 561

fol) and opioid (sufentanil) drug infusion,120 having drugs may play a role in subsequent psychological
potential implications for interruption of sedation in impairment in the ICU setting. Further, they should
susceptible patients. Importantly, clinicians must be alert for signs of emotional trauma and the
maintain a high index of suspicion for sedative and possible need for psychological interventions during
analgesic drug withdrawal as a patient recovers from recovery from critical illness.
critical illness because many of the symptoms, such
as restlessness, insomnia, delirium, nausea, hyper-
tension, tachycardia, diaphoresis, and fever, are non- Summary and Conclusions
specific and may be easily overlooked. Strategies for
gradual weaning of sedative and analgesic agents The majority of critically ill patients receive seda-
and/or substitution with orally or subcutaneously tive and/or analgesic medications to combat pain and
administered drugs in high-risk patients may be anxiety and to improve tolerance of the ICU envi-
useful.117,118,121 ronment. Patient-focused care related to these issues
includes assessment of predisposing and precipitat-
ing factors, frequent monitoring, careful medication
selection, and use of strategies to precisely target
Sedative and Analgesic Therapy and Long- therapy to defined end points and avoid sedation that
term Psychological Outcomes is excessive or prolonged.
The impact of sedative and analgesic medications
on neuropsychological health after recovery from
critical illness is a complex issue because of many References
confounding factors, such as interruption of seda- 1 Desbiens NA, Wu AW. Pain and suffering in seriously ill
tion, withdrawal from these agents or other sub- hospitalized patients. J Am Geriatr Soc 2000; 48:S183–S186
2 Sessler CN, Grap MJ, Brophy GM. Multidisciplinary man-
stances, critical illness and organ dysfunction, preex- agement of sedation and analgesia in critical care. Semin
isting neuropsychological disorders, physical restraint Respir Crit Care Med 2001; 22:211–225
and immobilization, pain, delirium, memory forma- 3 Novaes MA, Knobel E, Bork AM, et al. Stressors in ICU:
tion, and amnesia. Nevertheless, the influence of perception of the patient, relatives and health care team.
sedation on subsequent recall of real or imagined Intensive Care Med 1999; 25:1421–1426
4 Cooper AB, Thornley KS, Young GB, et al. Sleep in critically
events as well as the development of PTSD and ill patients requiring mechanical ventilation. Chest 2000;
depression is an important, yet underrecognized 117:809 – 818
issue. Nelson and colleagues122 observed a positive 5 Arroliga A, Frutos-Vivar F, Hall J, et al. Use of sedatives and
association between duration of sedation and depres- neuromuscular blockers in a cohort of patients receiving
sion-related and PTSD-related symptoms 6 to 41 mechanical ventilation. Chest 2005; 128:496 –506
6 Jones C, Backman C, Capuzzo M, et al. Precipitants of
months after acute lung injury. No difference in post-traumatic stress disorder following intensive care: a
PTSD was noted, however, between high and low hypothesis generating study of diversity in care. Intensive
sedative drug doses.122 Samuelson and coworkers123 Care Med 2007; 33:978 –985
found that patients who had no recall for ICU events 7 Magarey JM. Sedation of adult critically ill ventilated pa-
were more likely to have had heavy sedation with tients in intensive care units: a national survey. Aust Crit
Care 1997; 10:90 –93
fewer periods of wakefulness. Patients who subse- 8 Leape LL, Cullen DJ, Clapp MD, et al. Pharmacist partic-
quently had delusional memories had longer ICU ipation on physician rounds and adverse drug events in the
LOS, received higher doses of sedative drugs, and intensive care unit. JAMA 1999; 282:267–270
had more periods of being alert or agitated.123 9 Jacobi J, Fraser GL, Coursin DB, et al. Clinical practice
Interestingly, DIS was associated with less frequent guidelines for the sustained use of sedatives and analgesics
in the critically ill adult. Crit Care Med 2002; 30:119 –141
PTSD and a trend for better psychosocial adjustment 10 Ely EW, Shintani A, Truman B, et al. Delirium as a predictor
to illness when tested months later.114 These DIS of mortality in mechanically ventilated patients in the intensive
patients had shorter duration of MV and sedation, care unit. JAMA 2004; 291:1753–1762
yet had more episodes of being alert or agitated, than 11 Ouimet S, Kavanagh BP, Gottfried SB, et al. Incidence, risk
control patients. Some experts124 postulate that re- factors and consequences of ICU delirium. Intensive Care
Med 2007; 33:66 –73
call for factual events, rather than internally gener- 12 Peterson JF, Pun BT, Dittus RS, et al. Delirium and its
ated images such as hallucinations and nightmares, motoric subtypes: a study of 614 critically ill patients. J Am
may be protective for developing subsequent PTSD. Geriatr Soc 2006; 54:479 – 484
In recent work,6 these authors found associations 13 Pandharipande P, Jackson J, Ely EW. Delirium: acute
between prolonged sedation but also physical re- cognitive dysfunction in the critically ill. Curr Opin Crit
Care 2005; 11:360 –368
straint with little or no sedation and development of 14 Vaurio LE, Sands LP, Wang Y, et al. Postoperative delirium:
PTSD. Although further research is needed, clini- the importance of pain and pain management. Anesth Analg
cians should be aware that sedative and analgesia 2006; 102:1267–1273

562 Postgraduate Education Corner

 15 Pandharipande P, Shintani A, Peterson J, et al. Loraz- agents in critically ill patients. Crit Care Med 2006; 34:374 –
epam is an independent risk factor for transitioning to 380
delirium in intensive care unit patients. Anesthesiology 36 Botha J, Le Blanc V. The state of sedation in the nation:
2006; 104:21–26 results of an Australian survey. Crit Care Resusc 2005;
16 Mehta S, Granton J, MacDonald RJ, et al. High-frequency 7:92–96
oscillatory ventilation in adults: the Toronto experience. 37 Payen JF, Chanques G, Mantz J, et al. Current practices in
Chest 2004; 126:518 –527 sedation and analgesia for mechanically ventilated critically
17 Sessler CN. Sedation, analgesia, and neuromuscular block- ill patients: a prospective multicenter patient-based study.
ade for high-frequency oscillatory ventilation. Crit Care Anesthesiology 2007; 106:687– 695
Med 2005; 33:S209 –S216 38 Ramsay MA, Savege TM, Simpson BR, et al. Controlled
18 Gattinoni L, Tognoni G, Pesenti A, et al. Effect of prone sedation with alphaxalone-alphadolone. BMJ 1974; 2:656 –
positioning on the survival of patients with acute respiratory 659
failure. N Engl J Med 2001; 345:568 –573 39 Riker RR, Picard JT, Fraser GL. Prospective evaluation of
19 Murray MJ, Cowen J, DeBlock H, et al. Clinical practice the Sedation-Agitation Scale for adult critically ill patients.
guidelines for sustained neuromuscular blockade in the Crit Care Med 1999; 27:1325–1329
adult critically ill patient. Crit Care Med 2002; 30:142–156 40 Devlin JW, Boleski G, Mlynarek M, et al. Motor Activity
20 Kahn JM, Andersson L, Karir V, et al. Low tidal volume Assessment Scale: a valid and reliable sedation scale for use
ventilation does not increase sedation use in patients with with mechanically ventilated patients in an adult surgical
acute lung injury. Crit Care Med 2005; 33:766 –771 intensive care unit. Crit Care Med 1999; 27:1271–1275
21 Cheng IW, Eisner MD, Thompson BT, et al. Acute effects 41 Sessler CN, Gosnell MS, Grap MJ, et al. The Richmond
of tidal volume strategy on hemodynamics, fluid balance, Agitation-Sedation Scale: validity and reliability in adult
and sedation in acute lung injury. Crit Care Med 2005; intensive care unit patients. Am J Respir Crit Care Med
33:63–70 2002; 166:1338 –1344
22 Wolthuis EK, Veelo DP, Choi G, et al. Mechanical ventila- 42 De Jonghe B, Cook D, Griffith L, et al. Adaptation to the
tion with lower tidal volumes does not influence the pre- Intensive Care Environment (ATICE): development and
scription of opioids or sedatives. Crit Care 2007; 11:R77 validation of a new sedation assessment instrument. Crit
23 Nieszkowska A, Combes A, Luyt CE, et al. Impact of Care Med 2003; 31:2344 –2354
tracheotomy on sedative administration, sedation level, and 43 Weinert C, McFarland L. The state of intubated ICU
comfort of mechanically ventilated intensive care unit pa- patients: development of a two-dimensional sedation rating
tients. Crit Care Med 2005; 33:2527–2533 scale for critically ill adults. Chest 2004; 126:1883–1890
24 de Wit M, Wan SY, Gill S, et al. Prevalence and impact of 44 Ely EW, Truman B, Shintani A, et al. Monitoring sedation
alcohol and other drug use disorders on sedation and status over time in ICU patients: reliability and validity of
mechanical ventilation: a retrospective study. BMC Anes- the Richmond Agitation-Sedation Scale (RASS). JAMA
thesiol 2007; 7:3 2003; 289:2983–2991
25 Sessler CN. Sedation scales in the ICU. Chest 2004; 126: 45 Pun BT, Gordon SM, Peterson JF, et al. Large-scale imple-
1727–1730 mentation of sedation and delirium monitoring in the inten-
26 Chanques G, Jaber S, Barbotte E, et al. Impact of systematic sive care unit: a report from two medical centers. Crit Care
evaluation of pain and agitation in an intensive care unit. Crit Med 2005; 33:1199 –1205
Care Med 2006; 34:1691–1699 46 Weinert CR, Calvin AD. Epidemiology of sedation and
27 Sanders KD, McArdle P, Lang JD Jr. Pain in the intensive sedation adequacy for mechanically ventilated patients in a
care unit: recognition, measurement, management. Semin medical and surgical intensive care unit. Crit Care Med
Respir Crit Care Med 2001; 22:127–136 2007; 35:393– 401
28 Gelinas C, Fillion L, Puntillo KA, et al. Validation of the 47 De Jonghe B, Bastuji-Garin S, Fangio P, et al. Sedation
critical-care pain observation tool in adult patients. Am J algorithm in critically ill patients without acute brain injury.
Crit Care 2006; 15:420 – 427 Crit Care Med 2005; 33:120 –127
29 Payen JF, Bru O, Bosson JL, et al. Assessing pain in critically 48 Botha JA, Mudholkar P. The effect of a sedation scale on
ill sedated patients by using a behavioral pain scale. Crit ventilation hours, sedative, analgesic and inotropic use in an
Care Med 2001; 29:2258 –2263 intensive care unit. Crit Care Resusc 2004; 6:253–257
30 Blenkharn A, Faughnan S, Morgan A. Developing a pain 49 Costa J, Cabre L, Molina R, et al. Cost of ICU sedation:
assessment tool for use by nurses in an adult intensive care comparison of empirical and controlled sedation methods.
unit. Intensive Crit Care Nurs 2002; 18:332–341 Clin Intensive Care 1994; 5:17–21
31 Mateo OM, Krenzischek DA. A pilot study to assess the 50 LeBlanc JM, Dasta JF, Kane-Gill SL. Role of the bispectral
relationship between behavioral manifestations and self- index in sedation monitoring in the ICU. Ann Pharmacother
report of pain in postanesthesia care unit patients. J Post 2006; 40:490 –500
Anesth Nurs 1992; 7:15–21 51 Schneider G, Heglmeier S, Schneider J, et al. Patient State
32 Puntillo KA, White C, Morris AB, et al. Patients’ perceptions Index (PSI) measures depth of sedation in intensive care
and responses to procedural pain: results from Thunder patients. Intensive Care Med 2004; 30:213–216
Project II. Am J Crit Care2001; 10:238 –251 52 Cortinez LI, Delfino AE, Fuentes R, et al. Performance of
33 Puntillo KA, Morris AB, Thompson CL, et al. Pain behaviors the cerebral state index during increasing levels of propofol
observed during six common procedures: results from Thun- anesthesia: a comparison with the bispectral index. Anesth
der Project II. Crit Care Med 2004; 32:421– 427 Analg 2007; 104:605– 610
34 Gelinas C, Johnston C. Pain assessment in the critically ill 53 Kreuer S, Wilhelm W, Grundmann U, et al. Narcotrend
ventilated adult: validation of the Critical-Care Pain Obser- index versus bispectral index as electroencephalogram mea-
vation Tool and physiologic indicators. Clin J Pain 2007; sures of anesthetic drug effect during propofol anesthesia.
23:497–505 Anesth Analg 2004; 98:692– 697
35 Mehta S, Burry L, Fischer S, et al. Canadian survey of the 54 Vivien B, Di Maria S, Ouattara A, et al. Overestimation of
use of sedatives, analgesics, and neuromuscular blocking Bispectral Index in sedated intensive care unit patients

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 563

 revealed by administration of muscle relaxant. Anesthesiol- 76 Brook AD, Ahrens TS, Schaiff R, et al. Effect of a nursing-
ogy 2003; 99:9 –17 implemented sedation protocol on the duration of mechan-
55 Dasta JF, Kane SL, Gerlach AT, et al. Bispectral Index in ical ventilation. Crit Care Med 1999; 27:2609 –2615
the intensive care setting [letter]. Crit Care Med 2003; 77 Kress JP, Pohlman AS, O’Connor MF, et al. Daily interrup-
31:998 tion of sedative infusions in critically ill patients undergoing
56 Tonner PH, Wei C, Bein B, et al. Comparison of two mechanical ventilation. N Engl J Med 2000; 342:1471–1477
bispectral index algorithms in monitoring sedation in post- 78 Carson SS, Kress JP, Rodgers JE, et al. A randomized trial of
operative intensive care patients. Crit Care Med 2005; intermittent lorazepam versus propofol with daily interrup-
33:580 –584 tion in mechanically ventilated patients. Crit Care Med
57 Riker RR, Fraser GL. Monitoring sedation, agitation, anal- 2006; 34:1326 –1332
gesia, neuromuscular blockade, and delirium in adult ICU 79 Richman PS, Baram D, Varela M, et al. Sedation during
patients. Semin Respir Crit Care Med 2001; 22:189 –198 mechanical ventilation: a trial of benzodiazepine and opiate
58 Takamatsu I, Ozaki M, Kazama T. Entropy indices vs the in combination. Crit Care Med 2006; 34:1395–1401
bispectral index for estimating nociception during sevoflu- 80 Carrasco G, Cabre L, Sobrepere G, et al. Synergistic
rane anaesthesia. Br J Anaesth 2006; 96:620 – 626 sedation with propofol and midazolam in intensive care
59 Anderson RE, Barr G, Jakobsson JG. Correlation beween patients after coronary artery bypass grafting. Crit Care Med
AAI-index and the BIS-index during propofol hypnosis: a 1998; 26:844 – 851
clinical study. J Clin Monit Comput 2002; 17:325–329 81 Breen D, Karabinis A, Malbrain M, et al. Decreased dura-
60 Watson BD, Kane-Gill SL. Sedation assessment in critically tion of mechanical ventilation when comparing analgesia-
ill adults: 2001–2004 update. Ann Pharmacother 2004; based sedation using remifentanil with standard hypnotic-
38:1898 –1906 based sedation for up to 10 days in intensive care unit
61 Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechan- patients: a randomised trial. Crit Care 2005; 9:R200 –210
ically ventilated patients: validity and reliability of the confusion 82 Cigada M, Pezzi A, Di Mauro P, et al. Sedation in the
assessment method for the intensive care unit (CAM-ICU). critically ill ventilated patient: possible role of enteral drugs.
JAMA 2001; 286:2703–2710 Intensive Care Med 2005; 31:482– 486
62 Ely EW, Gautam S, Margolin R, et al. The impact of 83 Rhoney DH, Murry KR. National survey of the use of
delirium in the intensive care unit on hospital length of stay. sedating drugs, neuromuscular blocking agents, and reversal
Intensive Care Med 2001; 27:1892–1900 agents in the intensive care unit. J Intensive Care Med 2003;
63 Bergeron N, Dubois MJ, Dumont M, et al. Intensive Care 18:139 –145
Delirium Screening Checklist: evaluation of a new screening 84 Soliman HM, Melot C, Vincent JL. Sedative and analgesic
tool. Intensive Care Med 2001; 27:859 – 864 practice in the intensive care unit: the results of a European
64 Aldemir M, Ozen S, Kara IH, et al. Predisposing factors for survey. Br J Anaesth 2001; 87:186 –192
delirium in the surgical intensive care unit. Crit Care 2001; 85 Bauer TM, Ritz R, Haberthur C, et al. Prolonged sedation
5:265–270 due to accumulation of conjugated metabolites of midazo-
65 van der Mast RC, van den Broek WW, Fekkes D, et al. lam. Lancet 1995; 346:145–147
Delirium after cardiac surgery: the possible role of trypto- 86 McKenzie CA, McKinnon W, Naughton DP, et al. Differ-
phan in relation to the other neutral aminoacids. Adv Exp entiating midazolam over-sedation from neurological dam-
Med Biol 1996; 398:93–96 age in the intensive care unit. Crit Care 2005; 9:R32–36
66 Devlin JW, Fong JJ, Fraser GL, et al. Delirium assessment 87 Swart EL, van Schijndel RJ, van Loenen AC, et al. Contin-
in the critically ill. Intensive Care Med 2007; 33:929 –940 uous infusion of lorazepam versus medazolam in patients in
67 Hart RP, Levenson JL, Sessler CN, et al. Validation of a the intensive care unit: sedation with lorazepam is easier to
cognitive test for delirium in medical ICU patients. Psycho- manage and is more cost-effective. Crit Care Med 1999;
somatics 1996; 37:533–546 27:1461–1465
68 Hart RP, Best AM, Sessler CN, et al. Abbreviated cognitive 88 Pohlman AS, Simpson KP, Hall JB. Continuous intravenous
test for delirium. J Psychosom Res 1997; 43:417– 423 infusions of lorazepam versus midazolam for sedation during
69 Ouimet S, Riker R, Bergeon N, et al. Subsyndromal delir- mechanical ventilatory support: a prospective, randomized
ium in the ICU: evidence for a disease spectrum. Intensive study. Crit Care Med 1994; 22:1241–1247
Care Med 2007; 33:1007–1013 89 Barr J, Zomorodi K, Bertaccini EJ, et al. A double-blind,
70 Ely EW, Stephens RK, Jackson JC, et al. Current opinions randomized comparison of i.v. lorazepam versus midazolam
regarding the importance, diagnosis, and management of for sedation of ICU patients via a pharmacologic model.
delirium in the intensive care unit: a survey of 912 health- Anesthesiology 2001; 95:286 –298
care professionals. Crit Care Med 2004; 32:106 –112 90 Yaucher NE, Fish JT, Smith HW, et al. Propylene glycol-
71 Grap MJ, Borchers CT, Munro CL, et al. Actigraphy in the associated renal toxicity from lorazepam infusion. Pharma-
critically ill: correlation with activity, agitation, and sedation. cotherapy 2003; 23:1094 –1099
Am J Crit Care 2005; 14:52– 60 91 Walder B, Elia N, Henzi I, et al. A lack of evidence of
72 Chase JG, Agogue F, Starfinger C, et al. Quantifying superiority of propofol versus midazolam for sedation in me-
agitation in sedated ICU patients using digital imaging. chanically ventilated critically ill patients: a qualitative and
Comput Methods Programs Biomed 2004; 76:131–141 quantitative systematic review. Anesth Analg 2001; 92:975–983
73 Nilsestuen JO, Hargett KD. Using ventilator graphics to 92 Marik PE. Propofol: therapeutic indications and side-effects.
identify patient-ventilator asynchrony. Respir Care 2005; Curr Pharm Des 2004; 10:3639 –3649
50:202–234 93 Barrientos-Vega R, Mar Sanchez-Soria M, Morales-Garcia
74 Thille AW, Rodriguez P, Cabello B, et al. Patient-ventilator C, et al. Prolonged sedation of critically ill patients with
asynchrony during assisted mechanical ventilation. Intensive midazolam or propofol: impact on weaning and costs. Crit
Care Med 2006; 32:1515–1522 Care Med 1997; 25:33– 40
75 Kollef MH, Levy NT, Ahrens TS, et al. The use of contin- 94 Nimmo GR, Mackenzie SJ, Grant IS. Haemodynamic and
uous i.v. sedation is associated with prolongation of mechan- oxygen transport effects of propofol infusion in critically ill
ical ventilation. Chest 1998; 114:541–548 adults. Anaesthesia 1994; 49:485– 489

564 Postgraduate Education Corner

 95 Vasile B, Rasulo F, Candiani A, et al. The pathophysiology of need for ventilator support in a surgical intensive care unit.
propofol infusion syndrome: a simple name for a complex Qual Saf Health Care 2004; 13:203–205
syndrome. Intensive Care Med 2003; 29:1417–1425 110 Sessler CN. Wake up and breathe. Crit Care Med 2004;
96 Siobal MS, Kallet RH, Kivett VA, et al. Use of dexmedeto- 32:1413–1414
midine to facilitate extubation in surgical intensive-care-unit 111 Schweickert WD, Gehlbach BK, Pohlman AS, et al. Daily
patients who failed previous weaning attempts following interruption of sedative infusions and complications of crit-
prolonged mechanical ventilation: a pilot study. Respir Care ical illness in mechanically ventilated patients. Crit Care
2006; 51:492– 496 Med 2004; 32:1272–1276
97 Bamgbade OA. Dexmedetomidine for peri-operative seda- 112 Resar R, Pronovost P, Haraden C, et al. Using a bundle
approach to improve ventilator care processes and reduce
tion and analgesia in alcohol addiction. Anaesthesia 2006;
ventilator-associated pneumonia. Jt Comm J Qual Patient
61:299 –300
Saf 2005; 31:243–248
98 MacLaren R, Forrest LK, Kiser TH. Adjunctive dexmedeto-
113 Heffner JE. A wake-up call in the intensive care unit. N Engl
midine therapy in the intensive care unit: a retrospective J Med 2000; 342:1520 –1522
assessment of impact on sedative and analgesic require- 114 Kress JP, Gehlbach B, Lacy M, et al. The long-term
ments, levels of sedation and analgesia, and ventilatory and psychological effects of daily sedative interruption on criti-
hemodynamic parameters. Pharmacotherapy 2007; 27:351– cally ill patients. Am J Respir Crit Care Med 2003; 168:
359 1457–1461
99 Venn M, Newman J, Grounds M. A phase II study to 115 Kress JP, Vinayak AG, Levitt J, et al. Daily sedative inter-
evaluate the efficacy of dexmedetomidine for sedation in the ruption in mechanically ventilated patients at risk for coro-
medical intensive care unit. Intensive Care Med 2003; nary artery disease. Crit Care Med 2007; 35:365–371
29:201–207 116 Fonsmark L, Rasmussen YH, Carl P. Occurrence of with-
100 Dasta JF, Jacobi J, Sesti AM, et al. Addition of dexmedeto- drawal in critically ill sedated children. Crit Care Med 1999;
midine to standard sedation regimens after cardiac surgery: 27:196 –199
an outcomes analysis. Pharmacotherapy 2006; 26:798 – 805 117 Tobias JD. Tolerance, withdrawal, and physical dependency
101 Barr J, Donner A. Optimal intravenous dosing strategies for after long-term sedation and analgesia of children in the
sedatives and analgesics in the intensive care unit. Crit Care pediatric intensive care unit. Crit Care Med 2000; 28:2122–
Clin 1995; 11:827– 847 2132
102 Muellejans B, Lopez A, Cross MH, et al. Remifentanil 118 Berens RJ, Meyer MT, Mikhailov TA, et al. A prospective
versus fentanyl for analgesia based sedation to provide evaluation of opioid weaning in opioid-dependent pediatric
patient comfort in the intensive care unit: a randomized, critical care patients. Anesth Analg 2006; 102:1045–1050
double-blind controlled trial. Crit Care 2004; 8:R1–R11 119 Cammarano WB, Pittet JF, Weitz S, et al. Acute withdrawal
103 Fish DN. Treatment of delirium in the critically ill patient. syndrome related to the administration of analgesic and
Clin Pharm 1991; 10:456 – 466 sedative medications in adult intensive care unit patients.
104 Sharma ND, Rosman HS, Padhi ID, et al. Torsades de Crit Care Med 1998; 26:676 – 684
pointes associated with intravenous haloperidol in critically 120 Korak-Leiter M, Likar R, Oher M, et al. Withdrawal
ill patients. Am J Cardiol 1998; 81:238 –240 following sufentanil/propofol and sufentanil/midazolam se-
105 Milbrandt EB, Kersten A, Kong L, et al. Haloperidol use is dation in surgical ICU patients: correlation with central
associated with lower hospital mortality in mechanically nervous parameters and endogenous opioids. Intensive Care
ventilated patients. Crit Care Med 2005; 33:226 –229; dis- Med 2005; 31:380 –387
cussion 263–225 121 Tobias JD. Subcutaneous administration of fentanyl and
106 Skrobik YK, Bergeron N, Dumont M, et al. Olanzapine vs midazolam to prevent withdrawal after prolonged sedation
haloperidol: treating delirium in a critical care setting. in children. Crit Care Med 1999; 27:2262–2265
Intensive Care Med 2004; 30:444 – 449 122 Nelson BJ, Weinert CR, Bury CL, et al. Intensive care unit
107 MacLaren R, Plamondon JM, Ramsay KB, et al. A prospec- drug use and subsequent quality of life in acute lung injury
tive evaluation of empiric versus protocol-based sedation patients. Crit Care Med 2000; 28:3626 –3630
and analgesia. Pharmacotherapy 2000; 20:662– 672 123 Samuelson K, Lundberg D, Fridlund B. Memory in relation
108 Mascia MF, Koch M, Medicis JJ. Pharmacoeconomic impact to depth of sedation in adult mechanically ventilated inten-
of rational use guidelines on the provision of analgesia, sive care patients. Intensive Care Med 2006; 32:660 – 667
sedation, and neuromuscular blockade in critical care. Crit 124 Jones C, Griffiths RD, Humphris G, et al. Memory, delu-
Care Med 2000; 28:2300 –2306 sions, and the development of acute posttraumatic stress
109 Brattebo G, Hofoss D, Flaatten H, et al. Effect of a scoring disorder-related symptoms after intensive care. Crit Care
system and protocol for sedation on duration of patients’ Med 2001; 29:573–580

www.chestjournal.org CHEST / 133 / 2 / FEBRUARY, 2008 565