Professional Documents
Culture Documents
M aintaining an optimal level tion for a description of the methodology bined use of analgesics and sedatives may
of comfort and safety for used to develop these guidelines (2). ameliorate the stress response in critically
critically ill patients is a This document is limited to a discus- ill patients (7, 8). Pain may also contribute
universal goal for critical sion of prolonged sedation and analgesia. to pulmonary dysfunction through local-
care practitioners. The American College Consistent with the previous practice ized guarding of muscles around the area of
of Critical Care Medicine (ACCM) of the guidelines, this document pertains to pa- pain and a generalized muscle rigidity or
Society of Critical Care Medicine’s tients older than 12 years. The majority spasm that restricts movement of the chest
(SCCM’s) practice parameters for the op- of the discussion focuses on the care of wall and diaphragm (9). Effective analgesia
timal use of sedatives and analgesics was patients during mechanical ventilation. A may diminish pulmonary complications in
published in 1995 and recommended a discussion of regional techniques is not postoperative patients (10).
tiered approach to the use of sedatives included. Appendix A summarizes the Some patients recall unrelieved pain
and analgesics, largely on the basis of ex- recommendations made herein. when interviewed about their ICU stays (3,
pert opinion (1). These clinical practice 11, 12). The perception of pain can be in-
guidelines replace the previously published ANALGESIA fluenced by several factors, such as the ex-
parameters and include an evaluation of pectation of pain, prior pain experiences, a
the literature published since 1994 com- In these guidelines, “analgesia” is de- patient’s emotional state, and the cognitive
paring the use of these agents. The reader fined as the blunting or absence of sensa- processes of the patient (11). Patients
should refer to the accompanying introduc- tion of pain or noxious stimuli. Intensive should be educated about the potential for
care unit (ICU) patients commonly have pain and instructed to communicate their
pain and physical discomfort from obvious needs in an appropriate manner (such as
The American College of Critical Care Medicine factors, such as preexisting diseases, inva- using an assessment tool or other commu-
(ACCM), which honors individuals for their achieve- nication techniques). The goals of therapy
sive procedures, or trauma. Patient pain
ments and contributions to multidisciplinary critical
care medicine, is the consultative body of the Society and discomfort can also be caused by mon- should also be communicated to the pa-
of Critical Care Medicine (SCCM) that possesses rec- itoring and therapeutic devices (such as tient and family. In many cases, pain will be
ognized expertise in the practice of critical care. The catheters, drains, noninvasive ventilating managed but not completely eliminated.
College has developed administrative guidelines and devices, and endotracheal tubes), routine Fear of potent analgesics and misconcep-
clinical practice parameters for the critical care prac-
titioner. New guidelines and practice parameters are nursing care (such as airway suctioning, tions about pain and analgesics should be
continually developed, and current ones are system- physical therapy, dressing changes, and pa- addressed. Similarly, practitioner bias
atically reviewed and revised. tient mobilization), and prolonged immo- against the adequate use of opioids or mis-
Special thanks to E. Wesley Ely, MD, for his con- bility (3, 4). Unrelieved pain may contribute placed fears of adverse effects or addiction
tribution to the section on delirium. may produce inadequate prescribing or ad-
to inadequate sleep, possibly causing ex-
Address correspondence to Society of Critical Care
Medicine, 701 Lee Street, Suite 200, Des Plaines, IL haustion and disorientation. Agitation in an ministration (13, 14). Educating practitio-
60016. Available at www.sccm.org ICU patient may result from inadequate ners and assessing the quality of a pain
Key Words: analgesia; sedation; evidence-based pain relief. Unrelieved pain evokes a stress management program may improve anal-
medicine; fentanyl; hydromorphone; morphine; loraz- response characterized by tachycardia, in- gesia therapy, but such programs have not
epam; midazolam; propofol; haloperidol; guidelines
Copyright © 2002 by the Society of Critical Care creased myocardial oxygen consumption, been universally successful (4, 15). The im-
Medicine and the American Society of Health-System hypercoagulability, immunosuppression, portance of appropriate pain management
Pharmacists, Inc. and persistent catabolism (5, 6). The com- programs has been reinforced by the Joint
Equianalgesic
Agent Dose (i.v.) Half-life Metabolic Pathway Active Metabolites (Effect) Adverse Effects
Fentanyl 200 g 1.5–6 hr Oxidation No metabolite, parent accumulates Rigidity with high doses
Hydromorphone 1.5 mg 2–3 hr Glucuronidation None ...
Morphine 10 mg 3–7 hr Glucuronidation Yes (sedation, especially in renal Histamine release
insufficiency)
Meperidine 75–100 mg 3–4 hr Demethylation and Yes (neuroexcitation, especially in Avoid with MAOIsc and SSRIsd
hydroxylation renal insufficiency or high doses)
Codeine 120 mg 3 hr Demethylation and Yes (analgesia, sedation) Lacks potency, histamine release
glucuronidation
Remifentanil ... 3–10 min Plasma esterase None ...
Ketorolac ... 2.4–8.6 hr Renal None Risk of bleeding, GI and renal
adverse effects
a
More frequent doses may be needed for acute pain management in mechanically ventilated patients.
b
Cost based on 2001 average wholesale price.
c
MAOIs ⫽ monoamine oxidase inhibitors.
d
SSRIs ⫽ selective serotonin-reuptake inhibitors.
naloxone, is not recommended after pro- cause of their longer duration of effect. irritation with long-term use than tradi-
longed analgesia, because it can induce (Grade of recommendation ⫽ C) tional NSAIDs (56). The slow onset of
withdrawal and may cause nausea, cardiac action of some agents may decrease their
stress, and arrhythmias. Analgesics with Nonopioid Analgesics. NSAIDs provide utility for acute pain management.
agonist-antagonist action, such as nalbu- analgesia via the nonselective, competi- Acetaminophen is an analgesic used to
phine, butorphanol, and buprenorphine, tive inhibition of cyclooxygenase (COX), a treat mild to moderate pain. In combina-
can also elicit withdrawal symptoms and critical enzyme in the inflammatory cas- tion with an opioid, acetaminophen pro-
should be avoided during prolonged opioid cade. NSAIDs have the potential to cause duces a greater analgesic effect than higher
use. significant adverse effects, including gas- doses of the opioid alone (57). The role of
trointestinal bleeding, bleeding second- acetaminophen in critical care is limited to
Recommendations: A therapeutic plan ary to platelet inhibition, and the devel- relieving mild pain or discomfort, such as
and goal of analgesia should be estab- opment of renal insufficiency. Patients that associated with prolonged bed rest or
lished for each patient and communi- with hypovolemia or hypoperfusion, the use as an antipyretic. Care must be taken to
cated to all caregivers to ensure con- elderly, and those with preexisting renal avoid excessive and potentially hepatotoxic
sistent analgesic therapy. (Grade of impairment may be more susceptible to doses, especially in patients with depleted
recommendation ⫽ C) NSAID-induced renal injury (52, 53). Pro- glutathione stores resulting from hepatic
If intravenous doses of an opioid anal- longed use (more than five days) of ketoro- dysfunction or malnutrition. Acetamino-
gesic are required, fentanyl, hydro- lac has been associated with a two-fold in- phen should be maintained at less than 2 g
morphone, and morphine are the rec- crease in the risk of renal failure and an per day for patients with a significant his-
ommended agents. (Grade of increased risk of gastrointestinal and oper- tory of alcohol intake or poor nutritional
recommendation ⫽ C) ative-site bleeding (54, 55). NSAIDs should status and less than 4 g per day for others
Scheduled opioid doses or a continuous not be administered to patients with (Table 1) (58).
infusion is preferred over an “as needed” asthma and aspirin sensitivity.
Recommendations: NSAIDs or acet-
regimen to ensure consistent analgesia. Administration of NSAIDs may reduce
aminophen may be used as adjuncts to
A PCA device may be utilized to deliver opioid requirements, although the anal-
opioids in selected patients. (Grade of
opioids if the patient is able to under- gesic benefit of NSAIDs has not been sys-
recommendation ⫽ B)
stand and operate the device. (Grade of tematically studied in critically ill pa-
tients. Many oral agents are available, and Ketorolac therapy should be limited to a
recommendation ⫽ B)
ibuprofen and naproxen are available in maximum of five days, with close mon-
Fentanyl is preferred for a rapid onset of itoring for the development of renal in-
liquid form. Ketorolac is currently the
analgesia in acutely distressed patients. sufficiency or gastrointestinal bleeding.
only parenteral NSAID. The safety of ke-
(Grade of recommendation ⫽ C) Other NSAIDs may be used via the en-
torolac administration in patients with
Fentanyl or hydromorphone are pre- severe renal insufficiency or those under- teral route in appropriate patients.
ferred for patients with hemodynamic going dialysis has not been determined. (Grade of recommendation ⫽ B)
instability or renal insufficiency. The role, if any, of the more selective
(Grade of recommendation ⫽ C) COX-2 inhibitors in the critically ill re-
SEDATION
Morphine and hydromorphone are mains unknown. Selective COX-2 inhib- The indications for sedative agents are
preferred for intermittent therapy be- iting agents cause less gastrointestinal not well defined. Sedatives are common
consciousness and agitation from a sev- the assessment of sedation in adult criti- chanical ventilation. The desired level of
en-item list describing patient behavior cally ill patients (86). With the VICS scor- sedation should be defined at the start of
(Table 2). Excellent inter rater reliability ing system, patients are assessed inde- therapy and reevaluated on a regular ba-
has been demonstrated and validity has pendently for the ability to interact and sis as the clinical condition of the patient
been shown with two other scales. The communicate and for their level of activ- changes. Regimens should be written
Motor Activity Assessment Scale (MAAS), ity or restlessness. The VICS has not been with the appropriate flexibility to allow
adapted from the SAS, has also been val- tested to identify optimal sedation end- titration to the desired endpoint, antici-
idated and shown reliable for use in crit- points. Another scale, the Observer’s As- pating fluctuations in sedation require-
ically ill patients (83). The MAAS has sessment of Alertness/Sedation Scale, is ments throughout the day.
seven categories to describe patient be- often used in the operating room but Objective Assessment of Sedation. Ob-
haviors in response to stimulation (Table lacks the ability to assess agitation and jective testing of a patient’s level of seda-
2). The Ramsay scale measures three lev- has never been tested in the ICU (87). The tion may be helpful during very deep se-
els of awake states and three levels of COMFORT scale has been extensively dation or when therapeutic neuro-
asleep states (Table 2) (84). It has been tested and applied in the ICU environ- muscular blockade masks observable be-
shown to have an acceptable interrater ment, but only in children (88). havior. Vital signs, such as blood pressure
reliability compared with the SAS, but The appropriate target level of seda- and heart rate, are not specific or sensi-
has been criticized for its lack of clear tion will primarily depend on a patient’s tive markers of the level of sedation
discrimination and specific descriptors to acute disease process and any therapeutic among critically ill patients. Tools uti-
differentiate between the various levels and supportive interventions required. A lized in objective assessment include
(82, 85). Nevertheless, the Ramsay scale common target level of sedation in the heart rate variability and lower-esopha-
has been used in many comparative seda- ICU is a calm patient that can be easily geal contractility, but most are based on a
tion trials and is widely used clinically. aroused with maintenance of the normal patient’s electroencephalogram (EEG).
The Vancouver Interaction and Calmness sleep-wake cycle, but some may require The raw EEG signal has been manipu-
Scale (VICS) has also been validated for deep levels of sedation to facilitate me- lated by using several devices to simplify
bedside interpretation and improve reliabil- sedation. BIS has not been tested in pa- and are not yet proven useful in the ICU.
ity. For example, the bispectral index (BIS) tients with metabolic impairments or (Grade of recommendation ⫽ C)
uses a digital scale from 100 (completely structural abnormalities of the brain. Stud-
awake) to 0 (isoelectric EEG) (89). Most of ies have not compared the patient out- Sedation Therapy
the literature about the use of BIS in the comes of using BIS versus subjective scales.
operating room supports strong agreement Although BIS is likely to be useful when Benzodiazepines. Benzodiazepines are
between BIS and patient recall or level of patients are deeply comatose or under neu- sedatives and hypnotics that block the
hypnosis (90). Elective surgery patients re- romuscular blockade, routine use of this acquisition and encoding of new informa-
ceiving sedatives have shown a strong in- device cannot be recommended until the tion and potentially unpleasant experi-
verse correlation between hypnotic drug value and validity are confirmed. ences (anterograde amnesia) but do not
effect and BIS (91, 92). induce retrograde amnesia. Although
Although the BIS may be a promising Recommendations: A sedation goal or they lack any analgesic properties, they
tool for the objective assessment of seda- endpoint should be established and reg- have an opioid-sparing effect by moder-
tion or hypnotic drug effect, it has limita- ularly redefined for each patient. Regu- ating the anticipatory pain response (96,
tions in the ICU environment (93–95). BIS lar assessment and response to therapy 97). Benzodiazepines vary in their po-
scores may vary between patients at the should be systematically documented. tency, onset and duration of action, up-
same subjective level of sedation, and sub- (Grade of recommendation ⫽ C) take, distribution, metabolism, and pres-
jective scales may be more reproducible ence or absence of active metabolites
during light sedation (93, 94). Muscle- The use of a validated sedation assess- (Table 3). Patient-specific factors, such as
based electrical activity may artificially ele- ment scale (SAS, MAAS, or VICS) is age, concurrent pathology, prior alcohol
vate BIS scores if the patient has not re- recommended. (Grade of recommen- abuse, and concurrent drug therapy, af-
ceived neuromuscular blockade (94). A new dation ⫽ B) fect the intensity and duration of activity
version of BIS software is being tested for Objective measures of sedation, such as of benzodiazepines, requiring individual-
improved applicability in measuring ICU BIS, have not been completely evaluated ized titration. Elderly patients exhibit
Diazepam 2–5 min 20–120 hr Desmethylation and Yes (prolonged sedation) Phlebitis
hydroxylation
Lorazepam 5–20 min 8–15 hr Glucuronidation None Solvent-related acidosis/renal
failure in high doses
Midazolam 2–5 min 3–11 hr Oxidation Yes (prolonged sedation,
especially with renal
failure)
Propofol 1–2 min 26–32 hr Oxidation None Elevated triglycerides, pain
on injection
Haloperidol 3–20 min 18–54 hr Oxidation Yes (EPS)c QT interval prolongation
a
More frequent doses may be needed for management of acute agitation in mechanically ventilated patients.
b
Cost based on 2001 average wholesale price.
c
EPS ⫽ extrapyramidal symptoms.
slower clearance of benzodiazepines or rapid onset and awakening after single of 2 mg/mL every six hours) may lead to
their active metabolites and have a larger doses (Table 3) (78, 115). Because of its diarrhea because of the high PEG and PG
volume of drug distribution, contributing long-acting metabolites, a prolonged du- content (123).
to a marked prolongation of elimination ration of sedative effect may occur with Midazolam has a rapid onset and short
(111). Compromised hepatic or renal func- repeated doses, but this may be accept- duration with single doses, similar to di-
tion may slow the clearance of benzodiaz- able for long-term sedation (78). Loraz- azepam (Table 3) (115). The rapid onset
epines or their active metabolites. Induc- epam has a slower onset but fewer poten- of midazolam makes it preferable for
tion or inhibition of hepatic or intestinal tial drug interactions because of its treating acutely agitated patients. Accu-
enzyme activity can alter the oxidative me- metabolism via glucuronidation (Table 3) mulation and prolonged sedative effects
tabolism of most benzodiazepines (112). (98, 112). The slow onset makes loraz- have been reported in critically ill pa-
Benzodiazepine therapy should be ti- epam less useful for the treatment of tients using midazolam who are obese or
trated to a predefined endpoint, often re- acute agitation. Maintenance of sedation have a low albumin level or renal failure
quiring a series of loading doses. Hemo- can be accomplished with intermittent or (99 –103). Prolonged sedative effects may
dynamically unstable patients may continuous intravenous administration. also be caused by the accumulation of an
experience hypotension with the initia- Lorazepam has an elimination half-life of active metabolite, alpha-hydroxymidazo-
tion of sedation. Maintenance of sedation 12–15 hours, so an infusion is not readily lam, or its conjugated salt, especially in
with intermittent or “as needed” doses of titratable. Loading doses given by i.v. patients with renal insufficiency (101–
diazepam, lorazepam, or midazolam may push should be used initially with rela- 105). Significant inhibition of midazolam
be adequate to accomplish the goal of tively fixed infusion rates. Lorazepam in- metabolism has been reported with
sedation (78). However, patients requir- fusions should be prepared using the 2 propofol, diltiazem, macrolide antibiot-
ing frequent doses to maintain the de- mg/mL injection and diluted to a concen- ics, and other inhibitors of cytochrome
sired effect may benefit from a continu- tration of 1 mg/mL or less and mixed in a P450 isoenzyme 3A4, which could influ-
ous infusion by using the lowest effective glass bottle (116, 117). Despite these pre- ence the duration of effect (107, 108,
infusion dose. Continuous infusions cautions, precipitation may develop 112). Daily discontinuation of midazolam
must be used cautiously, as accumulation (118). An alternative is to administer un- infusions (wake up) with retitration to a
of the parent drug or its active metabo- diluted lorazepam as an infusion using a Ramsay scale endpoint reduced midazo-
lites may produce inadvertent overseda- PCA device (78). The lorazepam solvents lam requirements and was associated
tion. Frequent reassessment of a patient’s polyethylene glycol (PEG) and propylene with a reduction in the duration of me-
sedation requirements and active taper- glycol (PG) have been implicated as the chanical ventilation and length of ICU
ing of the infusion rate can prevent pro- cause of reversible acute tubular necro- stay (50). However, the patients in this
longed sedative effects (80). However, sis, lactic acidosis, and hyperosmolar trial were off of midazolam for an average
awakening times after several days of se- states after prolonged high-dose infu- of 5.3 hours per day, so this research
dation may be quite unpredictable in sions. The dosing threshold for this effect technique may be difficult to implement.
clinical use. In contrast, tolerance to ben- has not been prospectively defined, but Patients should be closely monitored for
zodiazepines may occur within hours to these case reports described doses that self-extubation or the removal of other
several days of therapy, and escalating exceeded 18 mg/hr and continued for monitoring devices during the daily
doses of midazolam have been reported longer than four weeks and higher doses awakening sessions.
(113, 114). While not well described in (⬎25 mg/hr) continuing for hours to days The routine use of a benzodiazepine
the literature, paradoxical agitation has (119 –121). It seems prudent to avoid doses antagonist, such as flumazenil, is not rec-
also been observed during light sedation of this magnitude. Alternatively, lorazepam ommended after prolonged benzodiaz-
and may be the result of drug-induced and diazepam may be administered via the epine therapy because of the risks of
amnesia or disorientation. enteral route in tablet or liquid form (122). inducing withdrawal symptoms and in-
Diazepam has been shown to provide Large doses of liquid lorazepam (i.e., 60 mg creasing myocardial oxygen consumption
Level of
Evidence Population Type
(Reference) (No. Patients) Exclusion Criteria Trial Design Drugs Mean Dosage
1 (150) MICU/SICU (100) Obese, head injury, Multicenter, Propofol Propofol: 1.77 mg/kg/hr
NMBA use open-label Midazolam Midazolam: 0.1 mg/kg/hr
Morphine: to all patients
2 (128) MICU/SICU (88 Neurologic injury, Open-label Propofol Propofol: 2.3 mg/kg/hr
total, 40 ongoing NMBA use Midazolam Midazolam: 0.17 mg/kg/hr
short-term Morphine: to all patients
sedation)
2 (151) CABG (30) Obese, renal or hepatic Open-label Propofol Propofol: 2.71 ⫾ 1.13 mg/kg/hr
insufficiency Midazolam Midazolam: 0.09 ⫾ 0.03 mg/kg/hr
Sufentanil: to all patients
1 (152) CABG (84) Renal or hepatic Open-label Propofol Propofol: 0.7 ⫾ 0.09 mg/kg/hr
insufficiency Midazolam Midazolam: 0.018 ⫾ 0.001 mg/kg/hr
Morphine: p.r.n.
1 (153) SICU (60) None Open-label, Propofol Propofol: 114.8 mg/hr (80.1 ⫾ 21.1
consecutive Midazolam kg)
patients Morphine: p.r.n. Midazolam: 2.1 ⫾ 1.3 mg/hr
(74.2 ⫾ 24 kg)
1 (154) MICU/SICU (61% Cardiac insufficiency, Multicenter, Lorazepam Lorazepam: 1.6 ⫾ 0.1 mg i.v. push,
trauma) neurosurgical or open-label Midazolam Midazolam: 14.4 ⫾ 1.2 mg infusion
unstable Morphine: p.r.n. over 8 hrs
1 (155) Cardiac surgery Renal, hepatic, or cardiac Double-blind Propofol Propofol: 0.64 ⫾ .17 mg/kg/hr
(41) failure Midazolam Midazolam: 0.015 ⫾ 0.001 mg/kg/hr
Morphine: p.r.n.
1 (156) CABG (75) Renal failure, neurologic Double-blind Propofol Propofol: 1.2 ⫾ 0.03 mg/kg/hr
history, hepatic failure, Midazolam Midazolam: 0.08 ⫾ 0.01 mg/kg/hr
cardiovascular Both Propofol and Propofol and Midazolam:
dysfunction Midazolam Propofol: 0.22 ⫾ 0.03 mg/kg/hr,
Morphine: to all patients Midazolam: 0.02 mg/kg/hr
2 (157) MICU/SICU (99) Neurosurgery, coma, Multicenter, Propofol Actual doses not specified
seizures open- Midazolam
label, Opioids: p.r.n.
intention-
to-treat
analysis
a
MICU ⫽ medical intensive care unit, SICU ⫽ surgical intensive care unit, NMBA ⫽ neuromuscular blocking agent, GCS ⫽ Glasgow Coma Score,
CABG ⫽ coronary artery bypass graft.
Sedative Selection Outcome is usually described in terms of als of sedation have compared propofol
the speed of onset, ability to maintain the and midazolam most often (eight of nine
Acute agitation arises from a variety of target level of sedation, adverse effects, trials) (Table 4). An opioid was available
etiologies, including pain. A short-acting
time required for awakening, and ability to all patients. Awakening times for pa-
opioid analgesic, such as fentanyl, may
to wean from mechanical ventilation. tients taking propofol ranged from 1 to
provide immediate sedation and patient
Most of the prospective, randomized tri- 105 minutes versus 1 to 405 minutes for
comfort; however, fentanyl has not been
als are experimentally flawed because patients receiving midazolam (128, 150 –
compared with other sedatives in a con-
trolled trial. Midazolam and diazepam they are unblinded, use uncontrolled 157). Time to extubation has also been
also have a rapid onset of sedation (115). amounts of opioids, and exclude patients compared, but other variables may influ-
Propofol has a rapid onset, but hypoten- with obesity or renal or hepatic insuffi- ence this outcome measure. Clinically,
sion and infusion-site pain can result ciency. This limits their general applica- these agents produced similar outcomes
from bolus dose administration. Cautious bility. There is a need for more large, following ⬍24 hours of infusion (Table 4).
use of sedatives is recommended for pa- high-quality, randomized trials of the ef- An intermediate duration of sedation
tients not yet intubated because of the fectiveness of different sedative agents (one to three days) was reported in three
risk of respiratory depression. (149). Most of the trials used a Ramsay randomized open-label trials (Table 5).
Comparative trials of prolonged seda- scale for assessment, so the depth of se- Propofol and midazolam were compared
tion have been performed in a variety of dation can generally be compared among for sedation of medical ICU patients and a
critical care settings. Many were sup- the trials. The trials are summarized in mixed medical-surgical ICU population
ported by pharmaceutical industry re- Tables 4 – 6. with respiratory failure (157, 158). Pa-
search grants; as a result, newer products Duration of Therapy. Short-term tients receiving propofol had statistically
have been evaluated more frequently. (⬍24 hours), randomized, open-label tri- more predictable awakening times than
Propofol: 20.2 hr Ramsay level 2–4 Awakening time: most awake at end of No statistical analysis reported
Midazolam: 21.3 hr infusion, longest: Propofol: 105 min,
Midazolam: 405 min
Propofol: 11.9 hr Ramsay level 2–5 and modified Time to extubation: Propofol: 0.3 hr p ⬍ 0.05, Propofol more rapid
Midazolam: 11.9 hr GCS Midazolam: 2.5 ⫾ 0.9 hr extubation
Propofol: 9.5 hr Ramsay level 5 Time to extubation: p ⬍ 0.01, Propofol more rapid
Midazolam: 9.8 hr Propofol: 250 ⫾ 135 min extubation
Midazolam: 391 ⫾ 128 min
Propofol: 9.2 hr Ramsay level 3 Time to extubation: Not statistically significant
Midazolam: 9.4 hr Propofol: 4.3 hr
Midazolam: 3.5 hr
16 hr observation, total Ramsay level 3 and response to Postsedation score at 5, 30, 60, and 90 p ⬍ 0.05, Midazolam more
sedation duration not stimulation min: Propofol scores lower at 5 and heavily sedated
defined 30 min
ⱕ24 hr—post hoc Ramsay scale, level was Time to extubation: p ⫽ 0.029 Overall: Propofol:
stratum specified daily Propofol: 5.6 hr 60.2% of time at target
Propofol: n ⫽ 21 Midazolam: 11.9 hr Ramsay score Midazolam:
Midazolam: n ⫽ 26 44% of time at target
Ramsay score, p ⬍ 0.05
a
MICU ⫽ medical intensive care unit, SICU ⫽ surgical intensive care unit, NMBA ⫽ neuromuscular blocking agent, GCS ⫽ Glasgow Coma Score,
CABG ⫽ coronary artery bypass graft.
patients receiving midazolam in both tri- Nine open-label, randomized trials com- midazolam were also compared for long-
als. Clinically, this time difference was paring long-term sedation (more than term sedation (159, 162). One of these
not as significant and did not produce three days) were reviewed in these guide- studies used a double-blind study design
more rapid discharge from the ICU (157). lines (Table 6) (70, 127–129, 157, 159 – (162). There was no statistically significant
In a three-way comparison of midazolam, 163). Most of the trials compared propofol difference in awakening time between these
lorazepam, and propofol infusions for se- with midazolam. All trials included opioid agents when titrated to similar levels of
dation of surgical ICU patients, the au- therapy, although administration was not sedation, although the awakening times as-
thors concluded that lorazepam was the controlled. Most studies used a Ramsay sociated with lorazepam appeared to be
preferred agent in this population (118). scale for patient assessment. In these trials, more predictable.
Overall, these agents were similar in the propofol consistently produced more rapid Sedative Comparison. Four trials
levels of sedation provided, the time re- awakening than midazolam with a statisti- compared lorazepam with midazolam
quired to achieve adequate sedation, and cal and, probably, a clinical difference (70, (118, 154, 159, 162). Intermittent loraz-
the number of dose adjustments per day. 127–129, 160, 161). Propofol patients awak- epam doses produced sedation compara-
However, midazolam produced adequate ened and were extubated in 0.25– 4 hours ble to a midazolam infusion during an
sedation during a greater percentage of while midazolam patients required 2.8 –10 eight-hour observation period (154).
time while propofol was associated with hours to awaken and up to 49 hours for Both lorazepam and midazolam have the
more undersedation and lorazepam with extubation (Table 6) (157). The greatest dif- potential to cause accumulation and pro-
more oversedation. Morphine was pro- ference in time to awakening was seen longed drug effects or oversedation if ad-
vided on an as needed basis and the av- when a deep level of sedation was the goal ministered excessively via continuous in-
erage dose was similar in all three of therapy (Ramsay level 4 –5). Patients re- fusion, especially when deep levels of
groups. Awakening times were not re- ceiving propofol awakened from deep seda- sedation are attempted (118, 159). How-
ported. Precipitation of lorazepam infu- tion significantly faster than those receiv- ever, a rigorous protocol of assessment
sions was reported (118). ing midazolam (127–129). Lorazepam and and titration of lorazepam infusions to
Level of
Evidence Population
(Reference) (No. Patients) Exclusion Criteria Design Drugs Mean Dosage
1 (158) MICU with respiratory ... Open-label Propofol Propofol: 1.25 ⫾ 0.87 mg/kg/hr
failure (73) Midazolam Midazolam: 3.1 ⫾ 3.2 mg/hr
Morphine: p.r.n.
2 (118) Surgical or Trauma Alcohol abuse, head Open-label Propofol Propofol: 2 ⫾ 1.5 mg/kg/hr
ICU (31) injury, dialysis Midazolam Midazolam: 0.04 ⫾ 0.03
Lorazepam mg/kg/hr
Morphine: p.r.n. Lorazepam: 0.02 ⫾ 0.01
mg/kg/hr
2 (157) MICU or SICU (99) Neurosurgery, coma, Multicenter, Open-label, Propofol Actual doses not specified
seizures intention-to-treat Midazolam
analysis Opioids: p.r.n.
a
MICU ⫽ medical intensive care unit, ICU ⫽ intensive care unit, SICU ⫽ surgical intensive care unit.
moderate levels of sedation produced a from 2.8 to 30 hours (Table 6). One center’s (bed availability) or patient-specific fac-
less variable awakening time with loraz- comparison of two concentrations of tors (other injuries and the need for ob-
epam than with midazolam, although the propofol (1% and 2%) versus midazolam servation) may impact a patient’s length
absolute difference in awakening time used in trauma patients has shown that of stay more than the sedation regimen.
was not statistically significant (159). A midazolam provides deep levels of sedation More rapid extubation with propofol was
nurse-managed sedation protocol, which more reliably than propofol, but the awak- not associated with a shorter length of
included the active titration of lorazepam ening times were much longer with mida- stay in a multi-center Canadian trial
infusions to a defined endpoint, avoided a zolam (127, 163). More patients receiving (157). Research that considers all of these
prolonged sedative effect compared with 1% propofol experienced failure because of cost factors is needed to estimate the
physician management of infusion rates elevated triglycerides, but the 2% propofol overall cost of sedative regimens. Since
(80). A blinded trial of lorazepam versus group experienced failure because of inad- the frequency of sedation-induced ad-
midazolam found that the lorazepam in- equate sedation. Failure of propofol to pro- verse effects has not been well described,
fusions were easier to manage than mi- vide adequate sedation of trauma patients there are insufficient data to create phar-
dazolam infusions because fewer dose ad- was reported elsewhere, although an expla- macoeconomic models comparing the
justments were required to maintain the nation was not apparent (118). potential costs of sedative regimens. The
desired level of sedation (162). In this Economic Comparison. Several of acquisition costs of sedatives vary widely
trial, wide inter- and intrapatient vari- these studies presented limited phar- among institutions, and costs may de-
ability was noted between sedative plasma macoeconomic data. In most reports, the cline with the availability of generic prod-
concentrations and the Ramsay score. No sedative costs were the only costs consid- ucts (Table 3).
difference was noted in patient recovery ered (cost minimization) (118, 154 –156, Multidisciplinary development and
when patients were evaluated for 24 162). Some studies included a portion of implementation of sedation guidelines
hours after the end of the infusion. Awak- the costs for ICU patient care (128, 161). have been shown to reduce direct drug
ening times were not reported in two of A sedative with a low acquisition cost may costs (from $81.54 to $18.12 per patient
the other trials (118, 154). be cited as the least expensive agent for per day), ventilator time (317 to 167
When titrated to a standard endpoint, prolonged sedation (e.g., lorazepam) (1, hours), and the lengths of ICU stay (19.1
midazolam and propofol provide compara- 32, 118). A complete economic analysis to 9.9 days) and total stay without a
ble levels of sedation with a similar onset of should consider costs associated with the change in mortality (165). Although an
effect (128, 150 –153, 155–157). As shown evaluation and treatment of sedation- economic analysis was not performed, a
in Table 4, there is generally no statistical induced adverse effects (e.g., prolonged nursing-implemented sedation protocol
or clinical difference in awakening times sedation, infection, and hypertriglyceri- using lorazepam reduced the duration of
between propofol and midazolam when demia), therapy failures (additional sedation and mechanical ventilation and
used for short-term sedation. Data from agents or high doses required), and drug the tracheostomy rate (80). A systematic
longer trials of sedation (more than 72 preparation and administration costs multidisciplinary team approach to seda-
hours) suggest that propofol is associated (precipitation and tubing changes) to de- tion and analgesia will produce clinical
with more reliable and rapid awakening, termine the total cost of therapy. The and economic benefits.
both statistically and clinically, than mida- cost of sedation-induced prolongation of An algorithm was developed to incor-
zolam (106, 127, 128, 156, 158, 160). Fol- ventilation or length of ICU stay is likely porate many of the assessment issues
lowing long-term sedation, propofol awak- to reduce the potential difference in ac- with the therapy options in this docu-
ening times ranged from 0.25 to 2.5 hours quisition costs between benzodiazepines ment (Figure 1). When using this algo-
and midazolam awakening times ranged and propofol (164). Institutional variables rithm, the pharmacology, potential ad-
Up to 3 days Behavior scale Awakening: defined by eye Many awake during infusion,
opening ability to follow with Propofol: smaller range of
eyes, hand grasp, and tongue awakening times, Used daily wake-
protrusion up to reassess patients
Propofol: 86.4 ⫾ 72 hr Ramsay level 2–4 Time with adequate sedation: Lorazepam vs. midazolam p ⫽ 0.03
Midazolam: 60 ⫾ 72 hr Midazolam: 79% Midazolam vs. Propofol p ⫽ 0.01
Lorazepam: 72 ⫾ 52.8 hr Propofol: 62% Propofol: more undersedation 31%
Lorazepam: 68% Lorazepam: more oversedation
14% and precipitation 18%
24–72 hr ⫽ post hoc stratum Ramsay scale, level was Time to extubation p ⫽ 0.068, Enrollment ended early,
Propofol: n ⫽ 21 specified daily Propofol: 7.4 hr insufficient power Overall:
Midazolam: n ⫽ 17 Midazolam: 31.3 hr Propofol: 60.2% of time at target
Ramsay score
Midazolam: 44% of time at target
Ramsay score, p ⬍ 0.05
verse effects, and therapeutic issues SEDATIVE AND ANALGESIC 169). Although not tested prospectively,
discussed in this document should be WITHDRAWAL it has been recommended that daily dose
considered. decrements of opioids not exceed 5–10%
Patients exposed to more than one in high-risk patients (170). If the drug is
Recommendations: Midazolam or di- week of high-dose opioid or sedative ther- administered intermittently, changing
azepam should be used for rapid seda- apy may develop neuroadaptation or the therapy to longer-acting agents may
tion of acutely agitated patients. physiological dependence. Rapid discon- also attenuate withdrawal symptoms
(Grade of recommendation ⫽ C) tinuation of these agents could lead to (171). Another recommendation for
withdrawal symptoms. Opioid withdrawal opioid weaning is to decrease a contin-
Propofol is the preferred sedative when
signs and symptoms include dilation of uous infusion rate by 20 – 40% initially
rapid awakening (e.g., for neurologic
the pupils, sweating, lacrimation, rhinor- and make additional reductions of 10%
assessment or extubation) is impor-
rhea, piloerection, tachycardia, vomiting, every 12–24 hours, depending on the
tant. (Grade of recommendation ⫽ B)
diarrhea, hypertension, yawning, fever, patient’s response (172). Conversion to
Midazolam is recommended for short- tachypnea, restlessness, irritability, in- a continuous subcutaneous infusion
term use only, as it produces unpre- creased sensitivity to pain, cramps, mus- has also been used for gradual fentanyl
dictable awakening and time to extu- cle aches, and anxiety. Benzodiazepine and midazolam weaning in children
bation when infusions continue longer withdrawal signs and symptoms include (173). Patient care costs may be in-
than 48 –72 hours. (Grade of recom- dysphoria, tremor, headache, nausea, creased unnecessarily if sedatives and
mendation ⫽ A) sweating, fatigue, anxiety, agitation, in- analgesics are withdrawn too slowly.
Lorazepam is recommended for the se- creased sensitivity to light and sound,
dation of most patients via intermit- paresthesias, muscle cramps, myoclonus, Recommendation: The potential for
tent i.v. administration or continuous sleep disturbances, delirium, and sei- opioid, benzodiazepine, and propofol
infusion. (Grade of recommenda- zures. Propofol withdrawal has not been withdrawal should be considered after
tion ⫽ B) well described but appears to resemble high doses or more than approxi-
benzodiazepine withdrawal. mately seven days of continuous ther-
The titration of the sedative dose to a
The occurrence of sedative and analge- apy. Doses should be tapered system-
defined endpoint is recommended with atically to prevent withdrawal
systematic tapering of the dose or sic withdrawal has been described in both
adult and pediatric ICU populations (166 – symptoms. (Grade of recommenda-
daily interruption with retitration to tion ⫽ B)
minimize prolonged sedative effects. 168). In adults, withdrawal is associated
(Grade of recommendation ⫽ A) with the length of stay, mechanical venti-
lation, and the dose and duration of anal- DELIRIUM
Triglyceride concentrations should be gesic and sedative therapy. Patients at high-
monitored after two days of propofol est risk include those who stay greater than As many as 80% of ICU patients have
infusion, and total caloric intake from seven days in the ICU, receive greater than delirium, characterized by an acutely
lipids should be included in the nutri- 35 mg/day of lorazepam, or greater than 5 changing or fluctuating mental status,
tion support prescription. (Grade of mg/day of fentanyl (166). inattention, disorganized thinking, and
recommendation ⫽ B) Studies of pediatric patients have an altered level of consciousness that may
The use of sedation guidelines, an algo- found that the rate of medication wean- or may not be accompanied by agitation.
rithm, or a protocol is recommended. ing may be very important in the devel- Placing severely ill patients in a stressful
(Grade of recommendation ⫽ B) opment of withdrawal syndromes (167, environment for prolonged periods exacer-
Level of
Evidence Population
(Reference) (No. Patients) Exclusion Criteria Design Drugs Mean Dosage
1 (129) MICU/SICU (108 Chronic liver disease, Open-label Propofol Propofol: 3.07–5.7 0.04 mg/kg/hr
consecutive) head injury, Midazolam Midazolam: 14 ⫾ 0.1 mg/kg/hr
ongoing NMBA use Morphine: to all patients
2 (70) MICU/SICU (68 None Open-label Propofol Propofol: 1.8 ⫾ 0.08 mg/kg/hr
consecutive) Midazolam Midazolam: 0.07 ⫾ 0.003 mg/kg/hr
Morphine: p.r.n.
2 (161) MICU/SICU (26) Hepatic or renal Open-label Propofol ⫹ alfentanil Alfentanil: 0.5–2 g/kg/hr
insufficiency, head Midazolam ⫹ morphine Propofol: 1–4 mg/kg/hr
injury, ongoing Morphine: 17–70 g/kg/hr
NMBA use Midazolam: 0.03–0.2 mg/kg/hr
1 (127) Trauma ICU (100 Renal or hepatic Open-label Propofol Propofol: 2.12 ⫾ 1.2 mg/kg/hr
consecutive) failure Midazolam Midazolam: 0.19 ⫾ 0.09 mg/kg/hr
Propofol ⫹ midazolam Propofol and Midazolam:
Morphine: p.r.n. Propofol: 1.6 ⫾ 0.05 mg/kg/hr,
Midazolam: 0.14 ⫾ 0.08 mg/kg/hr
1 (162) MICU (64) Head injury, ongoing Blinded Lorazepam Lorazepam: 23.1 ⫾ 14.4 mg/day
NMBA use Midazolam Midazolam: 372 ⫾ 256 mg/day
Fentanyl p.r.n.
2 (163) Trauma ICU (63 Renal or hepatic Open-label Propofol 2% Propofol: 6400 ⫾ 1797 mg/day
consecutive) failure Midazolam Midazolam: 297.8 ⫾ 103.8 mg/day
Morphine: to all patients
2 (157) MICU/SICU (99) Neurosurgery, coma, Multicenter, Propofol Actual doses not specified
seizures open-label, Midazolam
intention- Opioids: p.r.n.
to-treat
analysis
a
MICU ⫽ medical intensive care unit, ICU ⫽ intensive care unit, NMBA ⫽ neuromuscular blocking agent, GCS ⫽ Glasgow Coma Score; SICU, surgical
intensive care unit.
bates the clinical symptoms of delirium entation, and progressive confusion af- orated to develop and validate a rapid bed-
(174 –177). Delirium is usually character- ter sedative therapy. side instrument to accurately diagnose de-
ized by fluctuating levels of arousal lirium in ICU patients, who are often
throughout the day, associated with sleep- Assessment of Delirium nonverbal because they are on mechanical
wake cycle disruption, and hastened by ventilation. This instrument is called the
reversed day-night cycles (178). Delir- The gold standard criteria used to diag- Confusion Assessment Method for the
ium may be associated with confusion nose delirium is the clinical history and ICU (CAM-ICU) (181, 182). The work
and different motoric subtypes: hypoac- examination as guided by the Diagnostic was begun by Hart and colleagues with
tive, hyperactive, or mixed (179, 180). and Statistical Manual of Mental Disor- their publication of the Cognitive Test
Hypoactive delirium, which is associ- ders, 4th edition (DSM-IV) (177). Although for Delirium and a later version called
ated with the worst prognosis, is char- many scales and diagnostic instruments the Abbreviated Cognitive Test for De-
acterized by psychomotor retardation have been developed to facilitate the recog- lirium (183, 184). These two investiga-
manifested by a calm appearance, inat- nition and diagnosis of delirium, these tions were limited because they in-
tention, decreased mobility, and obtun- scales routinely exclude ICU patients be- cluded approximately 20 patients each
dation in extreme cases. Hyperactive cause it is often difficult to communicate and excluded some of the most severely
delirium is easily recognized by agita- with them (178, 181). Several groups of ill patients who are often cared for in
tion, combative behaviors, lack of ori- delirium investigators have recently collab- the ICU. These factors led the authors
Medium: Propofol: 116 hr Ramsay level 2–5 and modified Time to extubation: Medium: p ⬍ 0.05 Long: p ⬍ 0.05
Midazolam: 113 hr GCS Medium: Propofol: 0.4 ⫾ 0.1 hr
Long: Propofol: 312 hr Midazolam: 13.5 ⫾ 4 hr
Midazolam: 342 hr Long: Propofol: 0.8 ⫾ 0.3 hr
Midazolam: 36.6 ⫾ 6.8 hr
Lorazepam: 77 hr Ramsay level 2–3 Return to baseline mental status: Not statistically significant
Midazolam: 108 hr Lorazepam: 261 ⫾ 189 min,
Midazolam: 1815 ⫾ 2322 min
Propofol: 81 hr Midazolam: Own sedation scale Awakening lightly sedated: Light: p ⬍ 0.05 Heavy: p ⬍ 0.01
88 hr Propofol: 14 ⫾ 0.8 min
Midazolam: 64 ⫾ 20 min
Deep sedation: Propofol: 27 ⫾ 16 min
Midazolam: 237 ⫾ 222 min
Propofol: 139 hr Midazolam: Ramsay level 4–5 Time to t-tube Propofol: 4 ⫾ 3.9 hr p ⫽ 0.0001 Propofol: high
141 hr Midazolam: 48.9 ⫾ 47.2 hr triglycerides 12% men, 50%
women
Propofol: 99 hr Midazolam: Ramsay level 2–3 also rated Awakening: Propofol: 1.8 ⫾ 0.4 hr p ⬍ 0.02 More propofol patients
141 hr amnesia Midazolam: 2.8 ⫾ 0.4 hr with agitation upon awakening
Amnesia: Propofol: 29%
Midazolam: 100%
... Own scale, goal: moderate- Time to extubation: Propofol: 3 hr p ⫽ 0.006
heavy sedation (1–13 hr)
Midazolam: 50 hr (1–121 hr)
Propofol: 5.2 days Own scale, goal: moderate to Awakening (excluding head trauma): Midazolam vs Propofol or Propofol
Midazolam: 6.6 days heavy sedation Propofol: 110 ⫾ 50 min and Midazolam: p ⬍ 0.01;
Propofol and Midazolam: Midazolam: 660 ⫾ 400 min Propofol: high triglyceride levels
7.2 days Propofol and Midazolam: 190 ⫾ 200 min
Lorazepam: 141 hr Addenbrooke sedation scale, Awakening similar, times not reported Satisfactory sedation: Lorazepam:
Midazolam: 141 hr initial moderate to heavy 87 ⫾ 10.5%, Midazolam: 66.2 ⫾
sedation, tapering to light 23.1% p ⬍ 0.0001
sedation
Propofol: 6.5 hr Own scale, goal: moderate to Awakening (no head trauma) Awakening: not statistically
Midazolam: 11.1 hr heavy sedation Propofol: 145 ⫾ 50 min significant Less triglyceride
Midazolam: 372 ⫾ 491 min elevation than historical control
(reference 117)
24–72 hr—post hoc stratum Ramsay scale, level was Time to extubation: p ⫽ 0.03, enrollment ended early,
Propofol: n ⫽ 4 specified daily Propofol: 8.4 hr insufficient power limits
Midazolam: n ⫽ 10 Midazolam: 46.8 hr conclusion Overall: Propofol:
60.2% of time at target Ramsay
score Midazolam: 44% of time at
target Ramsay score, p ⬍ 0.05
to recommend that additional research demonstrated utility in important clinical sensitivity, and specificity. To complete
with delirium assessment tools be con- investigations (187, 188). the CAM-ICU, patients are observed for
ducted before routine application in Critical care nurses can complete de- the presence of an acute onset of mental
mechanically ventilated patients (184). lirium assessments with the CAM-ICU in status change or a fluctuating mental sta-
Collaboration among specialists in pul- an average of 2 minutes with an accuracy tus, inattention, disorganized thinking,
monary and critical care, neurology, psy- of 98%, compared with a full DSM-IV or an altered level of consciousness (Ta-
chiatry, neuropsychology, and geriatrics assessment by a geriatric psychiatric ex- ble 7). With the CAM-ICU, delirium was
has led to the development of a useful as- pert, which usually requires at least 30 diagnosed in 87% of the ICU patients
sessment tool (182, 185). It is based on the minutes to complete. The CAM-ICU as- with an average onset on the second day
Confusion Assessment Method (CAM), sessments have a likelihood ratio of over and a mean duration of 4.2 ⫾ 1.7 days
which was designed specifically for use by 50 for diagnosing delirium and high inter (185). Ongoing research will assist in un-
health care professionals without formal rater reliability (kappa ⫽ 0.96) (185). In derstanding the etiology of delirium and
psychiatric training, and incorporates the two subgroups expected to present the effects of therapeutic interventions.
DSM-IV criteria for the diagnosis of delir- the greatest challenge to the CAM-ICU Another instrument for delirium
ium (186). CAM, which is the most widely (i.e., those over 65 years and those with screening was validated in ICU patients
used delirium assessment instrument for suspected dementia), the instrument re- by comparison with a psychiatric evalua-
non-psychiatrists, is easy to use and has tained excellent inter rater reliability, tion (189). The use of these tools in pro-
1. Acute Onset of mental status changes or Is there evidence of an acute change in mental status from the baseline?
Fluctuating Course Did the (abnormal) behavior fluctuate during the past 24 hours, i.e., tend to come and go or
increase and decrease in severity?
Did the sedation scale (e.g., SAS or MAAS) or coma scale (GCS) fluctuate in the past 24 hours?a
2. Inattention Did the patient have difficulty focusing attention?
Is there a reduced ability to maintain and shift attention?
How does the patient score on the Attention Screening Examination (ASE)? (i.e., Visual
Component ASE tests the patient’s ability to pay attention via recall of 10 pictures; auditory
component ASE tests attention via having patient squeeze hands or nod whenever the letter
“A” is called in a random letter sequence)
3. Disorganized thinking If the patient is already extubated from the ventilator, determine whether or not the patient’s
thinking is disorganized or incoherent, such as rambling or irrelevant conversation, unclear
or illogical flow of ideas, or unpredictable switching from subject to subject.
For those still on the ventilator, can the patient answer the following 4 questions correctly?
1. Will a stone float on water?
2. Are there fish in the sea?
3. Does one pound weigh more than two pounds?
4. Can you use a hammer to pound a nail?
Was the patient able to follow questions and commands throughout the assessment?
1. “Are you having any unclear thinking?”
2. “Hold up this many fingers.” (examiner holds two fingers in front of patient)?
3. “Now do the same thing with the other hand.” (not repeating the number of fingers)
4. Altered level of consciousness (any level of Alert: normal, spontaneously fully aware of environment, interacts appropriately
consciousness other than alert (e.g., vigilant, Vigilant: hyperalert
lethargic, stupor, or coma) Lethargic: drowsy but easily aroused, unaware of some elements in the environment, or not
spontaneously interacting appropriately with the interviewer; becomes fully aware and
appropriately interactive when prodded minimally
Stupor: difficult to arouse, unaware of some or all elements in the environment, or not
spontaneously interacting with the interviewer; becomes incompletely aware and
inappropriately interactive when prodded strongly; can be aroused only by vigorous and
repeated stimuli and as soon as the stimulus ceases, stuporous subjects lapse back into the
unresponsive state.
Coma: unarousable, unaware of all elements in the environment, with no spontaneous
interaction or awareness of the interviewer, so that the interview is impossible even with
maximal prodding
Patients are diagnosed with delirium if they have both Features 1 and 2 and either Feature 3 or 4.
a
SAS ⫽ Sedation-Analgesia Scale, MAAS ⫽ Motor Activity Assessment Scale, GCS ⫽ Glasgow Coma Scale.
spective trials will delineate the long- causing a paradoxical increase in agita- chlorpromazine. Droperidol, a chemical
term ramifications of delirium on the tion (190). congener of haloperidol, is reported to be
clinical outcomes of ICU patients. The Neuroleptic agents (chlorpromazine more potent than haloperidol but has
study of delirium and other forms of cog- and haloperidol) are the most common been associated with frightening dreams
nitive impairment in mechanically venti- drugs used to treat patients with delir- and may have a higher risk of inducing
lated patients and other risk factors for ium. They are thought to exert a stabiliz- hypotension because of its direct vasodi-
neuropsychological sequelae after ICU ing effect on cerebral function by antag- lating and antiadrenergic effects (191,
care may be an important advancement onizing dopamine-mediated neuro- 192). Droperidol has not been studied in
in the monitoring and treatment of crit- transmission at the cerebral synapses and ICU patients as extensively as haloperidol.
ically ill patients. basal ganglia. This effect can also en- Haloperidol is commonly given via in-
Recommendation: Routine assessment hance extrapyramidal symptoms (EPS). termittent i.v. injection (193). The opti-
for the presence of delirium is recom- Abnormal symptomatology, such as hal- mal dose and regimen of haloperidol have
mended. (The CAM-ICU is a promising lucinations, delusions, and unstructured not been well defined. Haloperidol has a
tool for the assessment of delirium in thought patterns, is inhibited, but the long half-life (18 –54 hours) and loading
ICU patients.) (Grade of recommenda- patient’s interest in the environment is regimens are used to achieve a rapid re-
tion ⫽ B) diminished, producing a characteristic sponse in acutely delirious patients. A
flat cerebral affect. These agents also ex- loading regimen starting with a 2-mg
ert a sedative effect. dose, followed by repeated doses (double
Treatment of Delirium
Chlorpromazine is not routinely used the previous dose) every 15–20 minutes
Inappropriate drug regimens for seda- in critically ill patients because of its while agitation persists, has been de-
tion or analgesia may exacerbate delirium strong anticholinergic, sedative, and scribed (193, 194). High doses of haloper-
symptoms. Psychotic or delirious pa- ␣-adrenergic antagonist effects. Haloper- idol (⬎400 mg per day) have been re-
tients may become more obtunded and idol has a lesser sedative effect and a ported, but QT prolongation may result.
confused when treated with sedatives, lower risk of inducing hypotension than However, the safety of this regimen has