Journal of Critical Care 30 (2015) 344–347

Contents lists available at ScienceDirect

Journal of Critical Care

journal homepage: www.jccjournal.org

Guidelines

Effect of early mobilization on sedation practices in the neurosciences

intensive care unit: A preimplementation and
postimplementation evaluation☆
Robert Witcher, PharmD a, Lauren Stoerger, MSN, RN, CNRN, CCRN a, Amy L. Dzierba, PharmD, BCPS, FCCM a,
Amy Silverstein, MPH a, Axel Rosengart, MD, PhD b, Daniel Brodie, MD c, Karen Berger, PharmD, BCPS a,⁎
a
NewYork-Presbyterian Hospital, New York, NY
b
Weill Cornell Medical Center, New York, NY
c
Columbia University Medical Center, New York, NY

a r t i c l e i n f o a b s t r a c t

Keywords: Introduction: The use of sedation and analgesia protocols, daily interruption of sedation, and early mobilization
Early mobilization (EM) have been shown to decrease duration of mechanical ventilation and hospital length of stay (LOS).
Neurologic injury Methods: A retrospective chart review was conducted during a 6-month premobilization (pre-EM) and 6-month
Neurosurgery postmobilization (post-EM) period. Patients older than 18 years who were admitted to the neurosciences inten-
Critical care
sive care unit (ICU) and mechanically ventilated for at least 24 hours without documentation of withdrawal of life
Sedation
Analgesia
support or brain death were included.
Results: Thirty-one pre-EM and 37 post-EM patients were included. Baseline demographics were similar with the
exception of more ischemic stroke patients in the pre-EM group (P b .05). In the pre-EM and post-EM groups,
patients received similar cumulative doses of propofol, dexmedetomidine, and benzodiazepines but higher
median (interquartile range) doses of opioids (50.0 [13.8-165.0] vs 173.3 [41.2-463.2] μg of fentanyl equivalents
[P b .05]) in the post-EM group. Neurosciences ICU LOS was 10 (6-19) and 13 (8-18) days, respectively (P = .188).
Conclusions: After implementation of an EM program, an increase in opioid use and no significant change in other
sedatives were observed. Despite an increase in the amount of physical therapy and occupational therapy provided
to patients, there was no change in hospital and ICU LOS or duration of mechanical ventilation.
© 2014 Elsevier Inc. All rights reserved.

1. Introduction an increased number of ventilator-free days [9]. Patients' sedation was

Sedative and analgesic medications are used in critical care medicine
spanning a wide range of indications. Sedation practices have evolved
greatly from an era of heavy sedation to practices that aim to minimize
sedation. Landmark trials have demonstrated that daily sedation inter-
ruptions, spontaneous breathing trials, and protocolized sedation have
resulted in decreased duration of mechanical ventilation and length of
hospital and intensive care unit (ICU) stay [1-3]. With more patients
surviving acute illnesses, considerations for long-term implications
after ICU stays and outcomes such as delayed rehabilitation, irreversible
neurologic deficits, neurocognitive disability, and weakness become
prudent [4-8].
A previous trial in a medical and surgical ICU population demon-
strated that after the implementation of an early mobilization (EM)
protocol, patients experienced a decreased duration of delirium and
☆ Robert Witcher, Lauren Stoerger, Amy L. Dzierba, Amy Silverstein, Axel Rosengart,
Daniel Brodie, and Karen Berger declare that they have no conflicts of interest.
⁎ Corresponding author at: NewYork-Presbyterian Hospital/Weill Cornell Medical Center,
525 East 68th St, Room K-04, New York, NY 10065. Tel.: +1 212 746 0740.
E-mail address: kab9098@nyp.org (K. Berger).
managed using goal-directed sedation guided by the Richmond Assess-
ment for Sedation Scale and underwent daily interruption of sedative
agents. In this population, investigators found no change in sedative
and analgesic requirements. The only study to date to investigate the
effectiveness of increased mobility in a neurocritically ill population
was a prospective trial comparing a preintervention and postinter-
vention period after the implementation of a comprehensive mobility
initiative [10]. The initiative consisted of a progressive mobility
algorithm in a series of planned movements in a sequential manner
starting with the patient's current mobility status. The goal of the
initiative was to rehabilitate the patient to his or her functional base-
line. This program included activities of moving patients out of bed
and into a chair in the ICU and contained components to allow for
patient-activated progression through the algorithm, noting that
there are many physically functional neurointensive care patients,
such as subarachnoid hemorrhage patients. Patients included in this
study most frequently had a primary neurologic diagnosis of
subarachnoid hemorrhage followed by acute ischemic stroke, pres-
ence of an intracranial tumor, and intracerebral hemorrhage. Results
of this investigation revealed that after the implementation of the
initiative, the ICU and hospital length of stay (LOS) shortened

http://dx.doi.org/10.1016/j.jcrc.2014.12.003
0883-9441/© 2014 Elsevier Inc. All rights reserved.
 R. Witcher et al. / Journal of Critical Care 30 (2015) 344–347
significantly, as did the rates of hospital-acquired infection and
ventilator-associated pneumonia. Investigators did not detect a
difference in falls or inadvertent line disconnections between the 2
periods, and ultimately, the authors concluded that EM in neurologi-
cally injured patients was safe and effective in decreasing hospital
LOS and hospital-acquired infections.
Currently, the impact of an EM program on analgesia and sedation
patterns has not been evaluated in a neurocritically ill population.
Progressive practices including EM are not universally applied to
neurocritically ill patients due to fear of serious or sentinel adverse
events given some of the unique characteristics of this patient popula-
tion. However, subpopulations in the neurosciences ICU (NICU) may
benefit from practices of EM despite their neurologic injury. This study
sought to examine the effect of an EM protocol on sedation practices
of critically ill, mechanically ventilated patients in the NICU.
2. Patients and methods
We hypothesized that there would be a decrease in sedative and an-
algesic use as well as a decrease in duration of mechanical ventilation
and a decrease in hospital and ICU LOS after the implementation of an
EM program.
2.1. Setting
Our institution initiated an EM program beginning in January 2013
in an 11-bed NICU at a large, urban, academic medical center. At the
time of this study, there was no formalized protocol for sedation prac-
tices in this ICU, but Richmond Assessment for Sedation Scale scores
were monitored routinely and a clinical pharmacist was available dur-
ing weekdays to assess sedation for all patients. A dedicated team of
physical and occupational therapists assessed each patient in the NICU
daily for eligibility for mobilization using specific predefined criteria. Pa-
tients were considered ineligible for sedation interruption if they were
being managed for status epilepticus, treated with paralytics, diagnosed
with acute respiratory distress syndrome/hypoxemia, displayed evi-
dence of intracranial herniation, or had partial pressure of brain tissue
oxygen less than 15 mm Hg or intracranial pressure greater than
20 mm Hg. Patients who completed a sedation interruption were subse-
quently assessed for eligibility for mobilization. Patients ineligible for
EM included those with any contraindication for sedation interruption,
Glasgow Coma Scale (GCS) score less than or equal to 8 or coma off se-
dation, locked in syndrome, clinically symptomatic vasospasm, unstable
spine despite brace, cerebrospinal fluid leak, subdural hematoma evac-
uation within previous 24 hours, unsecured ruptured aneurysm, or
alteplase administration within previous 24 hours. The physical and oc-
cupational therapists rounded with the charge nurse daily and deter-
mined appropriate therapies for each patient. Interventions included
passive range of motion, active range of motion, bed-level activities in-
cluding sitting and leg dangling, and standing and ambulation. In coor-
dination with the primary team, physical therapy (PT) and occupational
therapy (OT) sessions were then prescribed and administered by the
team of therapists. Although pain was assessed before therapy was ad-
ministered, pretreatment with analgesics was not a routine component
of this program.
2.2. Patients
This was a retrospective chart review of patients who were admitted
to the NICU from January to June 2012 (pre-EM) and January to June
2013 (post-EM), 18 years or older, and intubated in the NICU and me-
chanically ventilated for at least 24 hours. Included patients could not
have documentation of withdrawal of life support or brain death.
345
2.3. Data collection
Demographic and clinical data were collected from the electronic
medical record, including age, sex, weight, medical history, NICU admit-
ting diagnosis, admission GCS score, and hospital and ICU LOS. The use
of therapeutic hypothermia, presence of vasopressors on admission,
and presence of a brain tissue oxygen monitor or external ventricular
drain were also collected. Daily cumulative doses of analgesic and seda-
tive agents were collected as well as the requirement of mechanical
ventilation and use of antipsychotic agents. The number of PT and OT
sessions administered was collected from the electronic medical record,
including pre-PT/OT and post-PT/OT pain scores and intensity of activi-
ties. Discharge disposition was also collected and recorded as to home,
to home with a health aide, or to a rehabilitation facility.
2.4. Outcomes
The primary outcome was the cumulative daily amount of analgesics
and sedatives used in the NICU before and after the implementation of
an EM program. The secondary outcomes included number of PT and
OT sessions before and after implementation of an EM protocol, change
in pain after PT/OT sessions, a description of the intensity of activity dur-
ing PT/OT, use of antipsychotics, duration of mechanical ventilation, ICU
and hospital LOS, and discharge disposition.
2.5. Statistical analysis
The statistical analyses were conducted using Minitab software
(version 16; Minitab, Inc, State College, PA) and SPSS Statistics (version
21; IBM, Armonk, NY). Discrete variables were expressed as percent-
ages, and continuous variables were expressed as median and inter-
quartile range, except when noted as mean and 95% confidence
interval. For comparison between pre-EM and post-EM groups, χ 2 or
2-sided t test was used for all categorical data, and Mann-Whitney U,
for all continuous variables' medians and interquartile ranges. P b .05
was considered statistically significant.
3. Results
A total of 180 patients were screened for inclusion, of which 112
were not included (55 mechanically ventilated b24 hours or in NICU
b24 hours, 30 documentation of withdrawal of life support or brain
death, 14 mechanically ventilated outside the NICU, and 13 intubated
or received a tracheostomy outside the NICU), resulting in 68 patients
included in this analysis: 31 in the pre-EM group and 37 in the post-
EM group. Baseline demographics are listed in Table 1. The median
age was 61 years; most common admission diagnoses were acute ische-
mic stroke, subarachnoid hemorrhage, and intracerebral hemorrhage;
and median admission GCS score was 7 in the pre-EM group and 8 in
the post-EM group. There were statistically significantly more patients
being treated for acute ischemic stroke in the pre-EM group (22.6% vs
5.4%, P = .049); however, all other characteristics were statistically sim-
ilar between groups. There was no statistical difference in the number of
patients requiring vasopressor therapy on admission to the NICU or use
of neuromuscular blocking agents, therapeutic hypothermia, external
ventricular drains, or brain tissue oxygen monitors during NICU stay.
There was a statistically significant increase in the cumulative opioid
dose during ICU stay in the pre-EM vs post-EM groups (50.0 μg/d [13.8-
165.0] vs 173 μg/d [41.2-463.2], P = .012). The cumulative doses of
benzodiazepines, propofol, and dexmedetomidine were not statistically
different between the 2 periods (Table 2). Statistically, more patients
received analgesic or sedative infusions after the implementation of
EM (80.6% vs 97.3%, P = .041).
After the implementation of EM, there was a statistically significant
increase in the mean number (95% confidence interval) of PT sessions
per patient (2.1 [1.6-2.7] vs 4.2 [3.6-4.9], P b .001) and OT sessions per
346 R. Witcher et al. / Journal of Critical Care 30 (2015) 344–347

Table 1 Table 3
Baseline characteristics Description of PT and OT sessions

Pre-EM Post-EM P Pre-EM Post-EM P

n = 31 n = 37 No. of PT or OT sessions 128 308 –

No. of PT sessions per patienta 2.1 (1.6-2.7) 4.2 (3.6-4.9) b.001
Age, median years (IQR) 61.3 (51.1-74.7) 60.6 (39.8-66.7) NS
No. of OT sessions per patienta 2.0 (1.6-2.6) 4.1 (3.5-4.8) b.001
Malea 36.4 54.3 NS
Change in pain after mobilization
Weight, kg 73.7 77.6 NS
Decrease in pain 1.6% 0.3% NS
Medical historya
No change in pain 96.1% 97.7% NS
Hypertension 41.9 43.2 NS
Increase in pain 2.3% 1.9% NS
Hyperlipidemia 35.5 29.7 NS
Intensity of therapy
Diabetes mellitus 19.4 13.5 NS
Passive range of motion 10.2% 5.2% NS
Atrial fibrillation 12.9 13.5 NS
Active range of motion 6.3% 2.3% NS
Prior stroke 6.5 2.7 NS
Sitting 64.8% 71.8% NS
Epilepsy 9.6 5.4 NS
Standing/walking 18.8% 20.8% NS
Admitting diagnosisa
a
Intracerebral hemorrhage 9.7 18.9 NS Values are presented as mean and 95% confidence interval.
Subarachnoid hemorrhage 12.9 16.2 NS
Acute ischemic stroke 22.6 5.4 .049
Subdural hematoma 9.7 2.7 NS in the pre-EM group. Notable nonsignificant trends include a greater
Status epilepticus 9.7 13.5 NS proportion of intracerebral, subarachnoid hemorrhage, and respiratory
Elective procedure 16.1 13.5 NS failure patients in the post-EM group, differences that could potentially
Seizure 16.1 8.1 NS
Encephalopathy 0.0 2.7 NS
reach significance in a larger patient population. Severity of illness
Respiratory failure 0.0 5.4 NS scores such as Acute Physiology and Chronic Health Evaluation II or
Other 3.2 13.6 – Simplified Acute Physiology Score II may not accurately reflect severity
Admission GCS, median (IQR) 7 (6-12) 8 (7,13) NS of illness in a neurocritically ill patient given the potential for an isolated
Therapeutic hypothermiaa 3.2 8.1 NS
neurologic injury; thus, the admission GCS score was used as a surrogate
Paralytic infusiona 3.2 10.8 NS
Vasopressor on admissiona 32.3 21.6 NS for severity of illness in this evaluation as well as use of therapeutic
Status epilepticus during NICUa 9.7 13.5 NS hypothermia, presence of vasopressors on admission, and presence of
Brain tissue oxygen monitor usea 12.9 10.8 NS a brain tissue oxygen monitor or external ventricular drain [11]. None
EVD usea 35.5 45.9 NS of these potential markers for severity of illness were statistically
IQR indicates interquartile range; NS, nonsignificant. different between groups.
a
Values are presented as percentages. In the present study, the increase in opioid use after the implemen-
tation of EM could theoretically be due to several factors. First, this
patient (2.0 [1.6-2.6] vs 4.1 [3.5-4.8], P b .001) provided to patients. An increase may be explained by increasing pain experienced by patients
increase in pain after PT or OT was experienced after 2.3% of sessions being mobilized, although in our study, there was no statistical differ-
in the pre-EM group vs 1.9% of patients in the post-EM group (P = ence in pain after PT or OT sessions. Several included disease states
0.726). The intensity of PT and OT sessions was statistically similar such as intracranial hemorrhages, notably those requiring craniotomies,
between groups (Table 3). can be painful, and EM of these patients may lead to increased analgesia
The proportion of patients managed with an antipsychotic agent requirements; however, this analysis did not include pain scales for
during their NICU stay was not statistically different between groups each disease state. Second, there was a statistically significant increase
(25.8% vs 12.5%, P = .230).The duration of mechanical ventilation was in the proportion of patients receiving infusions of analgesics and seda-
statistically similar between groups (5 [3-18] days vs 7 [4-19] days, tives; potentially, this increase in infusions translates to an increase in
P = .720), and the hospital LOS (22 [12-34] days vs 23 [16-35] days, cumulative opioid use when compared to a bolus dosing approach.
P = .416) and ICU LOS (10 [6-19] days vs 13 [8-18] days, P = .188) This analysis did not compare analgesic requirements of patients with
were statistically similar between groups (Fig. 1). There was no statisti- specific neurologic injuries. Furthermore, this evaluation did not com-
cally significant difference in the proportion of patients with a disposi- pare opioid requirements respective of the level of intensity of PT/OT
tion to home (9.7% vs 5.4%, P = .653) or disposition to home, to home sessions, as higher intensity PT/OT may yield increased analgesia
with a home health aide, or to a rehabilitation facility (54.8% vs 62.2%, requirements. Such analyses would result in small subpopulations
P = .624). yielding data with limited clinical tangibility. Interestingly, the use of
dexmedetomidine did increase, although nonsignificantly, which may
4. Discussion indicate an effort to incorporate lighter sedation to facilitate liberation
from the ventilator.
In this evaluation, we found that benzodiazepine, propofol, or
dexmedetomidine use was unchanged after the implementation of an
EM program, a finding previously observed in a medical and surgical
population; however, we found a statistically significant increase in opi-
oid use, a novel finding in critically ill patients [8]. The baseline charac-
teristics of the included population were statistically similar with the
exception of more patients being managed for an acute ischemic stroke

Table 2
Cumulative analgesic and sedative use

Pre-EM Post-EM P

Fentanyl equivalents (μg/d) 50.0 (13.8-165.0) 173.3 (41.2-463.2) .012

Midazolam equivalents (mg/d) 0.7 (0.0-8.38) 0.57 (0.0-5.37) NS
Propofol (mg/d) 334.0 (0.0-847.0) 250.2 (15.2-472.3) NS
Dexmedetomidine (μg/d) 75.6 (37.1-258.0) 134.4 (73.3-283.8) NS
Fig. 1. Duration of mechanical ventilation, ICU LOS, and hospital LOS.
 R. Witcher et al. / Journal of Critical Care 30 (2015) 344–347
Schweickert et al [9] evaluated the impact of an EM protocol with
goal-directed sedation on return to independent functional status in
mechanically ventilated critically ill patients. Patients' sedation was
interrupted daily, and subjects were randomized to receive early
exercise and mobilization during sedation holidays or to PT and OT as
prescribed by providers. The authors found that when mechanically
ventilated patients were mobilized earlier in their ICU stays, they
experienced more ventilator-free days and a shorter duration of deliri-
um. There was no statistical difference in the duration of sedative or
analgesic agent use or in the amount of propofol, opioids, or benzodiaz-
epines between groups. Notably, there were no unplanned extubations,
falls, or significant decreases in blood pressure, which is consistent
with previous studies demonstrating the feasibility and safety of mobi-
lization [6,9,12]. Similar to other trials, this population was exclusively
medically ill [8].
There are many unique characteristics of the neurocritically ill
population that must be taken into account when considering sedation
and mobilization practices [12,13]. Within this population, it is difficult
to determine if the patient's primary neurologic process rather than
iatrogenic causes or the clinical setting is precipitating agitation or dis-
comfort. Neurologically injured patients may also be instrumented
with external ventricular drains, intracranial monitors, or devices for
multimodality monitoring that can be physically cumbersome or irritat-
ing. Specific neurologic injuries require unique considerations as well.
For instance, patients with elevated intracranial pressures, who are
managed with paralytics, or who require sedative agents for the treat-
ment of status epilepticus would not be candidates for sedation holidays
or EM given their requirement for continuous sedative agents.
The trend toward increased hospital and ICU LOS is a surprising
finding that is inconsistent with prior literature, and the statistical
difference of fewer acute ischemic stroke patients in the pre-EM group
could contribute to this difference in LOS. Studies of acute ischemic
stroke patients reported a median LOS of 5 days [14,15]. Given this
information, one may expect to see a shorter LOS in the group
with more acute ischemic stroke patients, which was seen in the pre-
EM group.
As an additional explanation for the increased LOS, the EM program
in the present study was implemented in the absence of a sedation
guideline, a notable difference from a prior landmark study, in which
an EM protocol in critically ill patients was implemented concurrently
with goal-directed sedation [8]. This unimodal approach is not consis-
tent with current pain, agitation, and delirium guidelines from the
Society of Critical Care Medicine, which promote a bundled approach
to analgesia and sedation incorporating awakening and breathing trials,
judicious choice of sedative and analgesic agents, delirium assessments,
and EM concurrently with each other [16]. The trend of increased
hospital and ICU LOS may be partially explained by the unbundled
approach described in the implementation of this EM program, specifi-
cally implementation of a mobilization program in the absence of a
sedation guideline or protocol.
Several limitations can be described of the present study. This was a
single-center, retrospective analysis with a small sample population in
which many types of neurologic injuries were included. Many baseline
characteristics were statistically similar between groups, but notable
trends include more patients being managed for intracerebral hemor-
rhages, subarachnoid hemorrhages, and status epilepticus in the post-
EM group, possibly representing an unbalanced sample population. This
evaluation did not investigate the impact of the EM program on
sedation practices in specific neurologic injuries, all of which may be
associated with considerable differences in management. Potential con-
founders for increased duration of mechanical ventilation or LOS such as
development of a deep venous thrombosis, central line–associated blood-
stream infection, ventilator-associated pneumonia, and others were not
347
collected in this analysis. In addition, nurses in this institution were edu-
cated on recognition and management of inpatient pain control after the
implementation of the EM program, which may have increased identifica-
tion and subsequent treatment of pain in the post-EM period.
On a fundamental level, the population included in this evaluation is
heterogeneous by nature despite being limited to neurologic and neuro-
surgical illnesses, as demonstrated by the myriad admitting diagnoses.
These conditions are managed very differently, both medically and sur-
gically. Although the only statistically significant difference in admitting
diagnoses was acute ischemic stroke, there are several large trends that
may reach significance with a larger patient population. Given that the
neurocritically ill population is diverse, future research should aim to
evaluate the effects of EM on patients with specific neurologic injuries
to identify those who may experience maximal benefit from EM. Future
investigations should not only continue to focus on changes in analgesia
and sedation patterns but also address neuropsychiatric illness, func-
tional outcomes, and adverse events experienced with EM.
5. Conclusion
Implementation of an EM program in a heterogeneous neurocritically
ill population resulted in a significant increase in opioid use with no
significant change in benzodiazepine, propofol, or dexmedetomidine
use, potentially due to differences in disease states between populations.
Despite an increase in the number of PT and OT sessions provided to
patients, there was no statistical difference in duration of mechanical
ventilation or hospital or ICU LOS.
Acknowledgments
The authors thank Christine Kubin, Jennifer Michaels, and Kevin
McConeghy for their assistance with statistical analysis of these data.
References
[1] Kress JP, Pohlman AS, O’Connon MF, et al. Daily interruption of sedative infusions in
critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342:
1471–7.
[2] Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safety of a paired sedation and ven-
tilator weaning protocol for mechanically ventilated patients in intensive care: a
randomised controlled trial. Lancet 2008;371:126–34.
[3] Mehta S, Burry L, Cook D, et al. Daily sedation interruption in mechanically ventilat-
ed critically ill patients cared for with a sedation protocol: a randomized controlled
trial. JAMA 2012;308:1985–92.
[4] Cuthbertson BH, Roughton S, Jenkinson D, et al. Quality of life in the five years after
intensive care: a cohort study. Crit Care 2010;14:R6.
[5] Pandharipande PP, Girard TD, Jackson JC, et al. Long-term cognitive impairment after
critical illness. N Engl J Med 2013;369:1306–16.
[6] Herridge MS, Cheung AM, Tansey CM, et al. One-year outcomes in survivors of the
acute respiratory distress syndrome. N Engl J Med 2003;348:683–93.
[7] De Jonghe B, Lacherade JC, Durand MC, Sharshar T. Critical illness neuromuscular
syndromes. Crit Care Clin 2007;23:55–69.
[8] Schweickert WD, Hall J. ICU-acquired weakness. Chest 2007;131:1541–9.
[9] Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physical and occupational
therapy in mechanically ventilated, critically ill patients: a randomised controlled
trial. Lancet 2009;373:1874–82.
[10] Titsworth WL, Hester J, Correia T, et al. The effect of increased mobility on morbidity
in the neurointensive care unit. J Neurosurg 2012;116:1379–88.
[11] Hartley C, Cozens A, Mendelow AD, et al. The APACHE II scoring system in neurosur-
gical patients: a comparison with simple Glasgow coma scoring. Br J Neurosurg
1995;9:179–87.
[12] Bailey P, Thomsen GE, Spuhler VJ, et al. Early activity is feasible and safe in respira-
tory failure patients. Crit Care Med 2007;35:139–45.
[13] Stiller K. Safety issues that should be considered when mobilizing critically ill pa-
tients. Crit Care Clin 2007;23:35–53.
[14] Plakht Y, Pertov B, Gez H, et al. Stroke hospitalization over three decades: lower for men,
unchanged for women: a population-based study. J Women's Health 2014;23:1–6.
[15] Diringer MN, Edwards DF, Mattson DT, et al. Predictors of acute hospital costs for
treatment of ischemic stroke in an academic center. Stroke 1999;30:724–8.
[16] Barr J, Fraser GL, Puntillo K, et al. Clinical practice guidelines for the management of
pain, agitation, and delirium in adult patients in the intensive care unit. Crit Care
Med 2013;41:263–306.