Published October 22, 2024

NEJM Evid 2024;3(11)

DOI: 10.1056/EVIDra2300347

REVIEW

Sedation in the ICU

Kalynn A. Northam, Pharm.D., BCCCP,1 and Kristy M. Phillips, Pharm.D., BCCCP2

Abstract
Sedation practices are key to improving intensive care unit (ICU) outcomes. Adequate treat-
ment of pain, minimization of sedation, delirium prevention, and improved patient inter-
action to ensure early rehabilitation and faster ventilator liberation are evidenced-based
components of ICU care. Here we review components of appropriate ICU sedation includ-
ing the use of multicomponent care bundles such as the ABCDEF bundle with a focus on
changes in ICU practice that followed the Covid-19 pandemic.

Introduction

T
he coronavirus disease 2019 (Covid-19) pandemic altered practice across medi-
cine, but had a particularly lasting impact on critical care. During the pandemic,
intensive care units (ICUs) experienced reduced contact between patients and cli-
nicians, deeper sedation targets, increased utilization of neuromuscular blocking agents,
and reduced family visitation.1 In many ways these changes ran counter to what was con-
sidered good practice prior to the pandemic. The 2018 Clinical Practice Guidelines for the
Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep
Disruption in Adult Patients in the Intensive Care Unit (PADIS) laid out the key components
of optimal management of critically ill patients and were adopted as standard practice prior
to the pandemic.2 According to these guidelines, analgesic agents should be utilized first,
before sedatives, via a protocol that includes routine pain assessment with the Behavioral
Pain Scale (BPS, scored from 1 to 4 in each of the three domains of facial expression, upper
limb movement, and compliance with mechanical ventilation; higher scores indicate more
pain) or Critical Care Pain Observation Tool (CPOT, scored from 0 to 2 in each of the three
domains of facial expression, body movements, and muscle tension; higher scores indi-
cate more pain). In patients requiring sedative agents to facilitate mechanical ventilation,
clinicians should aim for light sedation using nonbenzodiazepine regimens with either
propofol or dexmedetomidine. Patients should be regularly evaluated for delirium using
the Confusion Assessment Method for the ICU (CAM-ICU) or the Intensive Care Delirium
Screening Checklist (ICDSC). Lastly, although considered a conditional recommendation
The author affiliations are listed at
based on low-quality evidence, multicomponent bundles should be implemented. Various the end of the article.
bundles have been reported in the literature and primarily focus on delirium prevention.3-6
The ABCDEF bundle is widely utilized and includes Assessment, prevention, and manage- Dr. Phillips can be contacted
ment of pain; Both spontaneous awakening trials (SATs) and spontaneous breathing trials at kristy​.phillips@dhha​.org or
at Denver Health & Hospital
(SBTs); Choice of analgesia and sedation; Delirium assessment, prevention, and manage- Authority, 777 Bannock Street,
ment; Early mobility and exercise, and; Family engagement and empowerment.2,7 Denver, CO 80204.
Since publication of the 2018 PADIS guidelines and the number of ventilator-free days as compared with morphine;
ABCDEF bundle, multiple studies have demonstrated however, mechanical ventilation across trial groups was less
positive impacts of bundle implementation on clinical out- than 2.5 days, limiting applicability to patients who require
comes, such as reductions in ICU length of stay and mortal- longer durations of support. In a retrospective cohort study,
ity, with an associated reduction in net health care costs.8-12 Choi and colleagues also evaluated ICU length of stay and
Additionally, in studies evaluating bundle compliance, even duration of mechanical ventilation among patients who
small improvements resulted in a dose-dependent reduc- received fentanyl- versus hydromorphone-based anal-
tion in delirium, increase in coma-free days, and increase gosedation.17 They identified no differences regarding ICU
in hospital survival.8,9 Most recently, Barr and colleagues length of stay or duration of mechanical ventilation. The
evaluated the aggregate effects of bundle implementa- fentanyl group had a higher use of dexmedetomidine and
tion on outcomes in mechanically ventilated patients and lower use of chemical paralysis, potentially highlighting
demonstrated reductions in ICU length of stay and days of differences in severity of illness between the two groups.
mechanical ventilation.8 Study hospitals had a high compli- These studies support the concept that any opioid can likely
ance rate for components “A” and “B” prior to the interven- be utilized if used in equianalgesic doses.
tion period, but improvements in “C” and “D” were made
Ketamine is an adjunctive pain option highlighted in the
with notably increased sedation assessments and delirium
PADIS guidelines, but data on its use are largely limited
screening followed by associated reductions in the use of
to postsurgical patients, with very low-quality evidence.1
benzodiazepine infusions. Despite this, emerging evidence
Ketamine has many favorable qualities including being less
since the pandemic indicates that overall bundle compli-
likely to cause hypotension, having analgesic properties,
ance may be reduced compared with prepandemic lev-
and potentially reducing opioid dosages.18 That cumula-
els.1,13 The focus of this narrative review is to highlight the
tive opioid doses may increase risk of delirium has led to
importance of evidenced-based sedation practices to ICU
an increasing body of research evaluating the use of ket-
outcomes, including the use of such multicomponent bun-
amine in conjunction with opioids or as monotherapy.16,19-24
dles. We review the components of the ABCDEF bundle
Multiple studies have demonstrated a reduction in sed-
and highlight literature that has been published since the
ative and opioid consumption and improved time within
2018 PADIS guidelines.
goal sedation parameters as measured by the Richmond
Agitation–Sedation Scale (RASS) of −1 to +1 with adjunc-
ASSESS, PREVENT, AND MANAGE PAIN tive ketamine use, although these outcomes are only sur-
The 2018 PADIS guidelines recommended utilizing anal- rogate outcomes.19,21,22 Ketamine use was associated with
gosedation (analgesia-first or analgesia-only strategies reduced rates of SAT and SBT implementation in another
ensure treatment of pain prior to the start of sedative study, leaving additional questions surrounding the impact
infusions which do not treat pain) via a protocol-based of ketamine use within a critically ill patient population.25
assessment and treatment plan.1 In a pooled analysis for
Even with a growing literature aimed at identifying the opti-
the guidelines, protocolized analgosedation reduced the
mal analgesic agent, there is currently not enough evidence
duration of mechanical ventilation and ICU length of stay,
to specify a first-line agent, and data regarding adjunctive
as well as reducing overall sedative requirements and
analgesia remain of low-quality. Clinicians should continue
reported pain intensity.2,14 The recent NONSEDA trial com-
to use clinical judgment, drug properties, and patient char-
pared an analgesia-only regimen with light sedation strate-
acteristics to develop analgosedation plans with titration
gies but found no significant impact on mortality between
using validated assessment tools. Additional research is
groups.15 Utilization of multimodal pharmacotherapy to
needed to determine if analgesic selection impacts short-
minimize opioid use is recommended by the 2018 guide-
and long-term outcomes and to determine optimal multi-
lines, although opioids remain the mainstay for pain man-
modal analgesia strategies.
agement. At the time of guideline publication, data were
lacking regarding optimal opioid selection.2 BOTH SPONTANEOUS AWAKENING TRIALS AND
To investigate differences in opioid choice, the ANALGESIC SPONTANEOUS BREATHING TRIALS
trial evaluated ventilator-free days at 28 days in patients SATs and SBTs have been shown to reduce the duration of
who received fentanyl- versus morphine-based analgose- mechanical ventilation and hospital length of stay.26 Both
dation.16 Use of fentanyl significantly increased the median SATs and SBTs should be performed daily in mechanically

NEJM EVIDENCE 2
ventilated patients meeting safety screen criteria. In with coma, irrespective of sedative choice. Mediation anal-
patients who do not pass SATs or SBTs, sedation should be ysis found that coma mediated 59% of in-hospital mortality
resumed at half the prior dose and re-titrated to achieve and, notably, the impact of Covid-19 diagnosis on mortal-
light sedation. Although daily sedation interruptions have ity was found to be insignificant.
become standard of care for mechanically ventilated
Despite the evidence supporting the use of light sedation,
patients, there remain barriers to consistent implemen-
clinicians may feel deeper sedation is required to control
tation. Balas and colleagues performed a secondary anal-
respiratory drive and minimize the potential for ventila-
ysis from the ICU Liberation Collaborative database and
tor-induced lung injury caused by asynchrony. Evidence
demonstrated that patients who had a documented seda-
for this approach is conflicting.31-33 Dzierba and colleagues
tion target and received more frequent sedation level
recently demonstrated that sedation depth was not cor-
assessments were more likely to have an SAT and SBT per-
related with respiratory drive at any evaluation time, high-
formed the next day.25 Additionally, patients receiving ket-
lighting the multitude of factors that impact respiratory
amine or benzodiazepines were less likely to have an SAT
drive outside of sedation.34 As an alternative to increasing
and SBT performed. Although a number of factors likely
sedation depth, investigators have evaluated the impact
impact successful implementation of SAT and SBT, this
of adjusting ventilator settings by increasing inspiratory
study demonstrates that sedation practices play a key role.
time or switching to pressure support ventilation when
Given the benefits of daily awakening and breathing trials,
feasible, which reduced the breath-stacking asynchrony
sedation protocols should focus on minimizing modifiable
index.35 Based on these data, a recent review proposed add-
factors that negatively impact implementation.
ing “R — control respiratory drive” to the ABCDEF bundle
to emphasize that ventilator settings should be addressed
CHOICE OF ANALGESIA AND SEDATION before utilizing deeper sedation and neuromuscular block-
The first step of selecting an analgesic and sedative agent is ing agents to avoid the negative outcomes associated with
setting a goal sedation depth. In a pooled analysis included deeper levels of sedation.36
in the PADIS guidelines, lighter sedation strategies were
Once clinicians select goal sedation depth, they should
associated with reduced time to extubation and rates of tra-
balance the risks and benefits of analgesic and sedative
cheostomy.2 A recent meta-analysis showed reduced dura-
agents, and utilize drug- and patient-specific parameters
tion of mechanical ventilation, time to extubation, ICU and
to select appropriate agents (Table 1). The PADIS guide-
hospital length of stay, and ventilator-associated pneumo-
lines recommend utilizing nonbenzodiazepines agents,
nia associated with light sedation.27 While the goal for most
either propofol or dexmedetomidine, as first line after
patients should be light sedation, the use of deeper seda-
the treatment of pain.1 Although not the primary focus of
tion levels may be clinically appropriate in patients who are
this review, the numerous landmark trials supporting this
receiving neuromuscular blockade, have refractory ventila-
recommendation are highlighted in Table 2.37-40 There is
tor asynchrony, or those receiving treatment for status epi-
no guideline consensus regarding the preferred nonben-
lepticus and refractory intracranial hypertension.
zodiazepine agent. The MENDS2 trial was a randomized
Utilizing deep sedation goals in patients with acute respi- trial that evaluated dexmedetomidine versus propofol in
ratory distress syndrome (ARDS) and Covid-19 offered achieving target sedation goals, and the impact on the
many lessons. The COVID-SED study evaluated the number of days alive without delirium or coma during
impact of early, deep sedation during the first 48 hours of the 14-day intervention period.41 Researchers found no
admission on in-hospital mortality.28 Investigators found difference in outcomes between these two agents when
early and deep sedation was associated with longer dura- titrated to light sedation. The SPICE III randomized trial
tion of mechanical ventilation and higher mortality among evaluated whether early use of dexmedetomidine ver-
patients with and without Covid-19, congruent to previ- sus usual care had an effect on all-cause mortality at 90
ously published data in a non–Covid-19 patient popula- days.42 This trial found no difference in mortality between
tion.29 Wongtangman and colleagues evaluated whether groups and increased risk of hypotension and bradycardia
patients with Covid-19 and ARDS were at a higher risk with dexmedetomidine use. Based on these trials, there is
of in-hospital mortality due to prolonged coma (defined currently no evidence to support one nonbenzodiazepine
as RASS ≤−3) than patients with non–Covid-19 ARDS.30 regimen over the other when targeting light levels of seda-
In-hospital mortality was significantly increased in patients tion. Clinicians should evaluate goal sedation depth and

NEJM EVIDENCE 3
 Table 1. Analgesia and Sedative Agent Key Characteristics.*

Onset,
Duration, Pharmacokinetics, Metabolism, Active
Agent Typical Dose Half-life Elimination Metabolite? Additional Considerations
Analgesia — Opioid:
Utilize bolus doses
before continuous
infusions
Fentanyl Intravenous Onset: 1–2 High first-pass clearance with No Chest wall rigidity/myoclonus
push: minutes approximately 75% excreted via reported with high doses or
25–50 μg every Duration: urine as inactive metabolites. when quickly administered.
30 minutes as 30–60 Hepatically metabolized Alterations in respiratory rate
needed minutes via CYP3A4, reduced or and alveolar ventilation may last
Infusion: increased elimination with longer than analgesic effect.
Infusion: context- CYP3A4 inhibitors or inducers,
25–200 μg/hr sensitive respectively; reduced elimination
half-life leads with heart failure, hepatic disease.
to increased Highly lipophilic with increased
duration of accumulation in patients receiving
effect with longer durations or in morbidly
prolonged obese patients.
infusion times
Hydromorphone Intravenous Onset: 5–10 Extensively metabolized via No May cause histamine release
push: minutes glucuronidation in the liver, with leading to hypotension, pruritus,
0.2–1 mg every Duration: 3–4 >95% metabolized to HM6G. urticaria or flushing.
2–3 hours as hours Patients with hepatic disease
needed may have increased exposure
but does not preclude them from
Infusion: use. Glucuronidation is generally
0.5–2 mg/hr preserved with hepatic disease.
Utilize lower initial starting doses
with slower uptitration.
Morphine Intravenous Onset: 5–10 Metabolized via glucuronidation Yes/No Increased likelihood of
push: 2–4 mg minutes in the liver to M3G and M6G. M3G: inactive histamine release leading to
every 1 hour as Duration: 2–4 Significantly altered in patients M6G: active urticaria, hives, hypotension.
needed hours with renal and hepatic impairment. — contributes M3G effects with accumulation:
M3G/M6G are primarily excreted significantly neurotoxicity and myoclonus;
Infusion: via the kidneys, prolonged half- to analgesic M6G effects with accumulation:
1–5 mg/hr life and accumulation with renal effect of significant sedation and
failure. morphine respiratory depression.
Analgesia —
Adjunctive
Ketamine Infusion Onset: 1 Complex and extensive hepatic Yes Emergence reactions may
analgesia minute metabolism: Norketamine occur, typically with initiation/
dosing: Offset: CYP3A4, CYP2B6, CYP2C19 discontinuation and with
approximately mediate N-demethylation into periprocedural bolus doses. Use
1–5 μg/kg/min 2 hours active metabolite norketamine. with caution in patients with
psychiatric disorders.
Use with caution in patients
with uncontrolled hypertension
or tachycardia.
Acetaminophen Oral/Rectal: Onset: Metabolized in the liver by No Scheduled doses may be more
650–1000 mg Oral: glucuronidation, sulfation, and effective at opioid sparing than
every 6 hours approximately oxidation via CYP2E1 to form a as-needed doses.
60 minutes reactive metabolite NAPQI which Use with caution in patients
Intravenous: Intravenous: is then metabolized to cysteine with acute hepatic injury,
500–1000 mg 15 minutes and mercapturic acid conjugates. cirrhosis, chronic alcohol use,
every 6 hours Half-life: 2.4 and chronic malnutrition
hours Intravenous acetaminophen has
a faster onset, but no current
evidence supports superiority
of intravenous acetaminophen
if patients can receive via
alternative routes.
(Continued)

NEJM EVIDENCE 4
 Table 1. (Continued ) Analgesia and Sedative Agent Key Characteristics.*

Onset,
Duration, Pharmacokinetics, Metabolism, Active
Agent Typical Dose Half-life Elimination Metabolite Additional Considerations
Nonsteroidal Generally not preferred in the ICU setting unless the patient is at low risk for complications
anti-inflammatory
drugs
Gabapentin Oral: 100–1200 Onset: 2–4 Eliminated via renal excretion as No Most beneficial in patients
mg every 8 hours unchanged drug experiencing neuropathic pain
hours Half-life: 5–7
Dose hours, directly
reductions proportional
required to creatinine
for renal clearance
dysfunction
Lidocaine Intravenous Onset: 1–2 Complex hepatic metabolism Yes Narrow therapeutic index,
load: 1–2 mg/ minutes via oxidative N-dealkylation by MEGX and GX early signs of toxicity typically
kg Offset: 20 CYP1A2 and CYP3A4 yielding neurologic in nature including
minutes metabolites MEGX and GX (both delirium, tremor, tinnitus, and
Infusion: Lidocaine active) as well as other inactive seizure. Severe toxicity may
1–2 mg/kg/hr demonstrates metabolites; cause bradycardia and heart
reduced Half-life is significantly prolonged block.
clearance in patients with hepatic Contraindicated in patients
and longer impairment; with Stokes–Adams syndrome,
half-life with With severe renal dysfunction, Wolff–Parkinson–White
prolonged lidocaine clearance is reduced and syndrome, or severe SA, AV, or
infusions accumulation of GX is increased. intraventricular block.
Sedatives — First line
Propofol Intravenous Onset: 1–2 Hepatic conjugation to inactive No May cause hypotension,
push: 10– minutes metabolites which are renally particularly with high doses
50 mg every Offset: excreted Provided as a lipid emulsion,
30 minutes as variable with monitor triglycerides at baseline
needed prolonged and every 48–72 hours
infusions Propofol-related infusion
Infusion: syndrome: diagnosis of
5–50 μg/kg/ exclusion, presents as
min severe metabolic acidosis,
hyperkalemia, lipemia,
rhabdomyolysis, renal failure,
EKG changes, or cardiac failure
Increased risk with prolonged
infusions or high doses
Dexmedetomidine Infusion: Onset: 30–60 Hepatic metabolism via No May cause bradycardia,
0.2–1.5 μg/ minutes glucuronidation and CYP2A6 hypotension, and drug fevers.
kg/hr Half-life: 2 aliphatic hydroxylation. Bolus doses may increase risk
hours Approximately 85% inactive for bradycardia.
metabolites are renally cleared Maintains light sedation with
within 24 hours. limited impact on respiratory
drive.
Patients can develop tolerance
and dependence with prolonged
infusions.
Sedatives —
Second Line
Ketamine Infusion Onset: 1 Complex and extensive hepatic Yes Emergence reactions may
sedative minute metabolism: Norketamine occur, typically with initiation/
dosing: Offset: CYP3A4, CYP2B6, CYP2C19 discontinuation and with
5–30 μg/kg/ approximately mediate N-demethylation into periprocedural bolus doses.
min 2 hours active metabolite norketamine. Use with caution in patient with
psychiatric disorders
Use with caution in patients
with uncontrolled hypertension
or tachycardia.
(Continued)

NEJM EVIDENCE 5
 Table 1. (Continued ) Analgesia and Sedative Agent Key Characteristics.*

Onset,
Duration, Pharmacokinetics, Metabolism, Active
Agent Typical Dose Half-life Elimination Metabolite Additional Considerations
Midazolam Intravenous Onset: Hepatic metabolism via CYP3A4 Yes Intermittent bolus doses
push: approximately to hydroxylated metabolites that α-hydroxymid- are preferred to continuous
2–4 mg every 3–5 minutes are conjugated and excreted azolam infusions.
30 min as Duration: renally Benzodiazepines may cause
needed 2–4 hours Highly lipophilic with increased paradoxical agitation and
(variable with accumulation in patients receiving delirium.
Infusion: prolonged longer durations or in morbidly
1–4 mg/hr infusions) obese patients
Lorazepam Intravenous Onset: 15–20 Hepatic metabolism via No Intermittent bolus doses
push: minutes conjugation into inactive are preferred to continuous
1–4 mg every Duration: 6–8 metabolites which are renally infusions.
30 minutes as hours cleared Benzodiazepines may cause
needed paradoxical agitation and
delirium.
Intravenous formulation
contains propylene glycol,
monitor for toxicity with
prolonged infusions including
anion gap metabolic acidosis.
* AV denotes atrioventricular; CYP1A2, cytochrome P450 1A2; CYP2A6, cytochrome P450 2A6; CYP2B6, cytochrome P450 2B6; CYP2C19, cytochrome
P450 2C19; CYP2E1, cytochrome P450 2E1; CYP3A4, cytochrome P450 3A4; EKG, electrocardiography; ICU, intensive care unit; M3G, morphine-
3-glucoronide; M6G, morphine-6-glucuronide; NAPQI, N-acetyl-p-benzoquinone imine; SA, sinoatrial; MEGX, monoethylglycinexylidide; GX,
glycinexylidide; and HM6G, hydromorphone-6-glucuronide.

potential adverse effects when choosing between propofol delirium.44 There was no difference in mean RASS score and
and dexmedetomidine. opioid and sedative total daily doses administered between
the two groups. Although additional research is needed,
ASSESS, PREVENT, AND MANAGE DELIRIUM results from this study may suggest that more consistent
sedation levels may reduce the incidence of delirium.
Sedative selection can affect the risk of developing ICU
delirium. Benzodiazepines should be avoided in favor of Practices during the Covid-19 pandemic provided addi-
alternative agents.2 Multicomponent bundles, such as tional insight into the impact of sedation on delirium.
ABCDEF, have been associated with a lower incidence of Bernard-Valnet and colleagues performed a retrospective
delirium.8 Beyond sedative agent selection and use of mul- observational study and evaluated incidence of delirium
ticomponent bundles, investigators have also evaluated the in patients with ARDS and Covid-19 versus another etiolo-
impact of other factors such as analgesic agent selection, gies.45 Interestingly, there was no difference in incidence of
RASS variability, sedative administration practices, and delirium between the Covid-19 and other etiology groups
adjunctive agents on preventing and treating delirium. despite higher total doses of propofol and fentanyl and
A secondary analysis of the ANALGESIC trial evaluated the longer durations of mechanical ventilation in the Covid-
incidence of delirium in patients who received fentanyl ver- 19 group. Use of sedative infusions was an independent
sus morphine for analgosedation.43 While the incidence of predictor of delirium. Additionally, in a multinational ret-
delirium was higher in the fentanyl group than in the mor- rospective observational study in critically ill patients with
phine group, the 95% confidence interval around that esti- Covid-19, delirium was found to be present in over half of
mate was wide and included the null. Interestingly, there the study population (54.9%) with a median duration of 3
was a linear association between the cumulative dose of days.24 Use of continuous benzodiazepine and opioid infu-
opioid received and the use of antipsychotics, suggesting sions was associated with delirium the next day. Results
that higher cumulative doses of opioids may increase the from these studies highlight the risk of delirium with ben-
likelihood of delirium. zodiazepines and also suggest that the use of other sedative
and opioid infusions may also play a role.
Ritchie and colleagues performed a retrospective observa-
tional study and demonstrated that mechanically ventilated Lastly, investigators have evaluated the utility of other
patients who developed delirium had a significantly higher medications such as antipsychotics and dexmedetomidine
coefficient of variation in RASS score than those without on treating and preventing delirium. In the MIND USA

NEJM EVIDENCE 6
 Table 2. Evidence for Specific Agents in Intensive Care Unit Sedation.*

Study Name Number of Time in Goal Delirium Time to ICU Length of

(Year) Study Medications Patients Sedation Prevalence Extubation Stay Mortality
Carson et al. Propofol versus 132 No difference N/A 5.8 days versus Survivors: 8.6 Hospital:
(2006)37 lorazepam 8.4 days days versus 25% versus
12.7 days 24%
MENDS Dexmedetomidine 107 Time in stated 79% versus N/A 7.5 days versus 28 day:
(2007)38 versus lorazepam goal 80% versus 82% 9 days 17% versus
67% 27%
SEDCOM Dexmedetomidine 366 Goal RASS −2 to 1 54% versus 3.7 days versus 5.9 days versus 30 day:
(2009)39 versus midazolam 77.3% versus 76.6% 5.6 days 7.6 days 22.5% versus
75.1% 25.4%
MIDEX Dexmedetomidine 500 Excluded deep 7.7% versus 4.2 days versus 8.8 days versus 45 day:
(2012)40 versus midazolam sedation 10% 6.1 days 10.1 days 27.3% versus
60.7% versus 21.1%
56.6%
PRODEX Dexmedetomidine 498 Excluded deep 4.9% versus 2.9 days versus 6.8 days versus 45 day:
(2012)40 versus propofol sedation 9.7% 3.9 days 7.7 days 17.1% versus
64.6% versus 19.4%
64.7%
SPICE III Dexmedetomidine 3904 Goal RASS score 40.7% versus 3 days versus 6 days versus 90 day:
(2019)42 versus usual −2 to 1 but deep 42.5% 3.3 days 6.3 days 29.1% versus
care (propofol sedation allowed 29.1%
or midazolam or as needed
combination) 56.6% versus
51.8%
MENDS2 Dexmedetomidine 432 Goal RASS score Days alive 3.8 days versus 6 days versus 90 day:
(2021)41 versus propofol −2 to 0 without 4.9 days 7 days 38% versus
57% versus 60% delirium 39%
10.7 versus
10.8 days
*ICU denotes intensive care unit; RASS, Richmond Agitation–Sedation Scale; and N/A, not applicable.

and AID ICU trials, antipsychotic agents administered to key aspect to including mechanically ventilated patients
patients with delirium did not impact the duration of delir- in these protocols was the use of light sedation in which
ium or days alive and out of the hospital at 90 days.46,47 Use patients were able to follow commands.50,54-56 Many investi-
of overnight dexmedetomidine in surgical patients has gators focused on initiating physical therapy once sedatives
demonstrated promising results, with some studies show- were discontinued.54,56 Conversely, other investigators
ing a reduction in incidence of delirium.48,49 allowed patients to participate in early mobility protocols
despite the need for sedatives, but coordinated these ses-
Although sedative agent selection and use of multicompo-
sions around daily awakening trials or used passive cycling
nent bundles both play important roles in preventing delir-
and range of motion.51,55 More recent trials evaluating the
ium, other factors such as analgesic agent selection, RASS
impact of early mobility have failed to demonstrate similar
variability, and use of continuous infusion sedatives and
positive outcomes.57-59 A trial in mechanically ventilated
opioids may also impact delirium outcomes. Additional
patients found no difference in days alive and out of the hos-
research is needed in this area to solidify these potential
pital at 180 days between the early mobility and standard
relationships.
care groups.57 There are many strengths of this trial, includ-
ing that it was a prospective randomized trial comparing an
EARLY MOBILITY AND EXERCISE individualized mobilization protocol to standard of care.
Results from initial early mobility and exercise studies One limitation is that authors state sedation was minimized
demonstrated improved functional status at hospital dis- to facilitate early mobilization, but do not provide specific
charge, shorter hospital and ICU length of stay, decreased details regarding weaning strategies. A high percentage of
ICU delirium days, increased ventilator-free days, and patients in the early mobility and usual care groups initially
lower odds of readmission or death.50-54 Investigators suc- received continuous infusion sedatives and were deeply
cessfully and safely implemented early mobility proto- sedated (median RASS score −3 [interquartile range −4 to
cols despite concurrent mechanical ventilation.50,52-58 One −2] vs. −3 [interquartile range −4 to −2], respectively). In

NEJM EVIDENCE 7
the early mobility group, sedation was the primary barrier or family participates in daily rounds or a family conference
preventing participation in out-of-bed activities up to day 5. is held.8,9 Implementation of this component was likely
Additionally, over 60% of the study population had a lowest impacted by visitation restrictions during the Covid-19
daily RASS score below −2 up to day 4. Although patients pandemic, but family members should be reincorporated
in the early mobility group had significantly longer mobi- into care given the positive impact on outcomes such as
lization sessions and were able to achieve higher levels of delirium, ICU length of stay, and caregiver anxiety and
mobilization sooner, sedation practices may have still influ- depression.64-67
enced the results. Kho and colleagues performed a random-
Limited evidence is available regarding family involvement
ized controlled trial in mechanically ventilated patients and
and perceptions of sedation practices in mechanically ven-
compared in-bed cycling with usual physical therapy versus
tilated patients. Prime and colleagues performed a pro-
usual physical therapy alone.58 There was no difference in
spective observational study and interviewed 16 awake,
Physical Function Test for ICU-scored between groups 3
mechanically ventilated patients and their family mem-
days after ICU discharge. Patients were allowed to partic-
bers to better understand their experience during light
ipate even if they were receiving continuous sedative infu-
sedation.68 The questionnaire evaluated eight categories
sions, and the trial protocol did not specify requirements
including preferences to be more awake. Fifty percent of
for daily awakening trials. Cycle effort could be passive or
patients agreed or strongly agreed that they preferred to
was considered active if patient cycling rates exceeded the
be kept awake, and the mean score between patients and
set ergometer speed with greater than zero watts of power
family members for this category was not significantly dif-
for at least 2 minutes. More than 50% of precycling RASS
ferent. Beyond patient and family preference, there may be
scores were −5 to −2, which may have contributed to the
additional long-term benefits to the use of lighter sedation
low percentage of active cycling days (29.1%). It may be
in regards to development of post-traumatic stress disorder
beneficial to integrate daily awakening trials with mobility
in ICU survivors.69
activities to increase active patient involvement.
Family members may also play a vital role in evaluating
There are many barriers that influence patients’ abil-
pain in critically ill patients. Puntillo and colleagues per-
ity to participate in early mobility. Institutional barriers
formed a prospective observational study of agreement
may include staffing shortages, lack of appropriate equip-
in pain score assessments for noncomatose critically ill
ment, and limited physical and occupational therapy
patients.70 Agreement in pain intensity score between the
resources.60,61 Kamdar and colleagues performed a second-
patient and family was the highest, followed by the physi-
ary analysis of a prospective observational study and exam-
cian, and the nurse. The CPOT is one of the gold standard
ined how patient-specific barriers impacted participation
pain assessment scales used for patients who are unable to
in daily physical therapy.62 They demonstrated a negative
communicate.1 Mohand-Saïd and colleagues performed
association between use of sedative infusions and opioid
a prospective observational study in which they enrolled
boluses with physical therapy participation. The use of opi-
10 family members to evaluate pain using the CPOT after
oid boluses may be a marker of uncontrolled pain driving
15-minute training sessions.71 Overall, all family members
lack of participation rather than a direct medication effect.
agreed or strongly agreed that they felt comfortable utilizing
Additionally, these results demonstrate use of sedatives
the CPOT and wanted to be more involved in pain assess-
may drive lack of participation instead of mechanical venti-
ments. Shahid and colleagues developed the CPOT-Fam in
lation. Similarly, Thomsen and colleagues performed a pro-
an effort to make the traditional CPOT more user friendly
spective pre- and postimplementation study and reported
for family members in which the core components were
absence of sedative use was an independent predictor of
reformatted into questions.72 Accuracy of the CPOT-Fam
increased ambulation in patients admitted to a respiratory
was assessed prospectively by having recruited participants
ICU.63 These studies support that limiting sedation can
and investigators score sample cases, which demonstrated
improve patient participation in early mobility and exercise.
high agreement between evaluators.
Family members can play a critical role in optimizing anal-
FAMILY ENGAGEMENT AND EMPOWERMENT gesia and sedation practices in the ICU. Family presence
The F component of the ABCDEF bundle was introduced to may allow for patients to tolerate lighter sedation, and fam-
ensure family engagement and empowerment was incorpo- ily members can assist with pain assessments and manage-
rated into the comprehensive care of patients in the ICU.8 ment for patients who are mechanically ventilated or unable
Compliance with this component is achieved if the patient to communicate. Investigators have evaluated risk factors

NEJM EVIDENCE 8
for development of post-intensive care syndrome in family management, and early exercise/mobility bundle*. Crit Care Med
members, and demonstrated various patient, relative, and 2014;42:1024-1036. DOI: 10.1097/CCM.0000000000000129.

medical staff-related factors likely contribute.73 Involving 4. Brummel NE, Girard TD, Ely EW, et al. Feasibility and safety of
family members in analgesic and sedation decisions may early combined cognitive and physical therapy for critically ill med-
help alleviate some of the medical staff-related factors such ical and surgical patients: the Activity and Cognitive Therapy in
as communication and perceptions of care. Future studies ICU (ACT-ICU) trial. Intensive Care Med 2014;40:370-379. DOI:
10.1007/s00134-013-3136-0.
are needed to better understand how to optimize the avail-
able sedation and pain scoring tools to incorporate family 5. Moon KJ, Lee SM. The effects of a tailored intensive care unit delir-
feedback and involvement. ium prevention protocol: a randomized controlled trial. Int J Nurs
Stud 2015;52:1423-1432. DOI: 10.1016/j.ijnurstu.2015.04.021.

6. Rivosecchi RM, Kane-Gill SL, Svec S, Campbell S, Smithburger

PL. The implementation of a nonpharmacologic protocol to pre-
Conclusions vent intensive care delirium. J Crit Care 2016;31:206-211. DOI:
Sedation practices are closely linked to ICU outcomes. 10.1016/j.jcrc.2015.09.031.

Implementation of evidence-based strategies such as pro- 7. Marra A, Ely EW, Pandharipande PP, Patel MB. The ABCDEF
tocolized analgosedation and minimization of sedation bundle in critical care. Crit Care Clin 2017;33:225-243. DOI:
to facilitate coordination of the ICU Liberation bundle 10.1016/j.ccc.2016.12.005.
should be promoted for all critically ill patients, as even 8. Pun BT, Balas MC, Barnes-Daly MA, et al. Caring for critically ill
small improvements in ventilator bundle compliance have patients with the ABCDEF bundle: results of the ICU liberation
demonstrated improvements in patient outcomes includ- collaborative in over 15,000 adults. Crit Care Med 2019;47:3-14.
ing duration of mechanical ventilation, ICU length of DOI: 10.1097/CCM.0000000000003482.
stay, delirium- and coma-free days, and hospital survival. 9. Barnes-Daly MA, Phillips G, Ely EW. Improving hospital survival
Moving past the Covid-19 pandemic, hospitals should con- and reducing brain dysfunction at seven California community
duct evaluations of bundle compliance and develop strate- hospitals: implementing PAD guidelines via the ABCDEF bundle
gies for continued process improvement, especially as they in 6,064 patients*. Crit Care Med 2017;45:171-178. DOI: 10.1097/

relate to analgesia- and sedative-based bundle components. CCM.0000000000002149.

10. Barr J, Downs B, Ferrell K, et al. Improving outcomes in mechan-

ically ventilated adult ICU patients following implementation
Disclosures
of the ICU liberation (ABCDEF) bundle across a large health-
Author disclosures and other supplementary materials are available at care system. Crit Care Explor 2024;6:e1001. DOI: 10.1097/
evidence​.nejm​.org. CCE.0000000000001001.

11. Hsieh SJ, Otusanya O, Gershengorn HB, et al. Staged implemen-

tation of awakening and breathing, coordination, delirium moni-
Author Affiliations
toring and management, and early mobilization bundle improves
1
Department of Pharmacy, Massachusetts General Hospital, 55 Fruit
Street, Boston, MA patient outcomes and reduces hospital costs*. Crit Care Med
2019;47:885-893. DOI: 10.1097/CCM.0000000000003765.
2
Department of Pharmacy, Denver Health and Hospital Authority,
777 Bannock Street, Denver, CO 12. Otusanya OT, Hsieh SJ, Gong MN, Gershengorn HB. Impact of
ABCDE bundle implementation in the intensive care unit on spe-
References cific patient costs. J Intensive Care Med 2022;37:833-841. DOI:
10.1177/08850666211031813.
1. Bakare LS, Groth CM, Young M, Hogan B, Rappaport SH. Com-
parison of intensive care unit liberation bundle adherence prior to 13. Liu K, Nakamura K, Katsukawa H, et al. ABCDEF bundle and sup-
and during the COVID -19 pandemic: a retrospective cohort study. portive ICU practices for patients with coronavirus disease 2019
JACCP J Am Coll Clin Pharm 2023;6:1313-1320. DOI: 10.1002/ infection: an international point prevalence study. Crit Care Explor
jac5.1878. 2021;3:e0353. DOI: 10.1097/CCE.0000000000000353.

2. Devlin JW, Skrobik Y, Gélinas C, et al. Clinical practice guide- 14. Strøm T, Martinussen T, Toft P. A protocol of no sedation for
lines for the prevention and management of pain, agitation/seda- critically ill patients receiving mechanical ventilation: a ran-
tion, delirium, immobility, and sleep disruption in adult patients domised trial. Lancet 2010;375:475-480. DOI: 10.1016/S0140-
in the ICU. Crit Care Med 2018;46:e825-e873. DOI: 10.1097/ 6736(09)62072-9.
CCM.0000000000003299.
15. Olsen HT, Nedergaard HK, Strøm T, et al. Nonsedation or light
3. Balas MC, Vasilevskis EE, Olsen KM, et al. Effectiveness and safety sedation in critically ill, mechanically ventilated patients. N Engl
of the awakening and breathing coordination, delirium monitoring/ J Med 2020;382:1103-1111. DOI: 10.1056/NEJMoa1906759.

NEJM EVIDENCE 9
16. Casamento AJ, Serpa Neto A, Young M, et al. A phase II clus- 28. Stephens RJ, Evans EM, Pajor MJ, et al. A dual-center cohort study
ter-crossover randomized trial of fentanyl versus morphine for on the association between early deep sedation and clinical out-
analgosedation in mechanically ventilated patients. Am J Respir comes in mechanically ventilated patients during the COVID-19
Crit Care Med 2021;204:1286-1294. DOI: 10.1164/rccm.202106- pandemic: the COVID-SED study. Crit Care 2022;26:179. DOI:
1515OC. 10.1186/s13054-022-04042-9.

17. Choi H, Radparvar S, Aitken SL, Altshuler J. Analgosedation: the 29. Shehabi Y, Bellomo R, Reade MC, et al. Early intensive care seda-
use of fentanyl compared to hydromorphone. J Crit Care Med tion predicts long-term mortality in ventilated critically ill patients.
2021;7:192-198. DOI: 10.2478/jccm-2021-0026. Am J Respir Crit Care Med 2012;186:724-731. DOI: 10.1164/
rccm.201203-0522OC.
18. Guillou N, Tanguy M, Seguin P, Branger B, Campion JP, Mallédant
Y. The effects of small-dose ketamine on morphine consump- 30. Wongtangman K, Santer P, Wachtendorf LJ, et al. Association of
tion in surgical intensive care unit patients after major abdomi- sedation, coma, and in-hospital mortality in mechanically venti-
nal surgery. Anesth Analg 2003;97:843-847. DOI: 10.1213/01. lated patients with coronavirus disease 2019–related acute respira-
ANE.0000075837.67275.36. tory distress syndrome: a retrospective cohort study*. Crit Care Med
2021;49:1524-1534. DOI: 10.1097/CCM.0000000000005053.
19. Garber PM, Droege CA, Carter KE, Harger NJ, Mueller EW.
Continuous infusion ketamine for adjunctive analgosedation in 31. de Wit M, Pedram S, Best AM, Epstein SK. Observational study
mechanically ventilated, critically ill patients. Pharmacother J of patient-ventilator asynchrony and relationship to sedation level.
Hum Pharmacol Drug Ther 2019;39:288-296. DOI: 10.1002/ J Crit Care 2009;24:74-80. DOI: 10.1016/j.jcrc.2008.08.011.
phar.2223.
32. Vaschetto R, Cammarota G, Colombo D, et al. Effects of propofol on
20. Amer M, Maghrabi K, Bawazeer M, et al. Adjunctive ketamine patient-ventilator synchrony and interaction during pressure sup-
for sedation in critically ill mechanically ventilated patients: an port ventilation and neurally adjusted ventilatory assist*. Crit Care
active-controlled, pilot, feasibility clinical trial. J Intensive Care Med 2014;42:74-82. DOI: 10.1097/CCM.0b013e31829e53dc.
2021;9:54. DOI: 10.1186/s40560-021-00569-1. 33. Asynchronies in the Intensive Care Unit (ASYNICU) Group, de
21. Magrum B, Elefritz JL, Eisinger G, McLaughlin E, Doepker B. Effi- Haro C, Magrans R, et al. Effects of sedatives and opioids on trig-
cacy of continuous infusion ketamine for analgosedation in the ger and cycling asynchronies throughout mechanical ventilation:
medical intensive care unit: a propensity-weighted analysis. J Pharm an observational study in a large dataset from critically ill patients.
Pract 2023;37:862-870. DOI: 10.1177/08971900231191154. Crit Care 2019;23:245. DOI: 10.1186/s13054-019-2531-5.

22. Chan K, Burry LD, Tse C, Wunsch H, De Castro C, Williamson DR. 34. Dzierba AL, Khalil AM, Derry KL, Madahar P, Beitler JR. Dis-
Impact of ketamine on analgosedative consumption in critically ill cordance between respiratory drive and sedation depth in criti-
patients: a systematic review and meta-analysis. Ann Pharmaco- cally ill patients receiving mechanical ventilation. Crit Care Med
ther 2022;56:1139-1158. DOI: 10.1177/10600280211069617. 2021;49:2090-2101. DOI: 10.1097/CCM.0000000000005113.

23. Wu TT, Ko S, Kooken R, Van Den Boogaard M, Devlin JW. Explor- 35. Chanques G, Kress JP, Pohlman A, et al. Impact of ventilator

ing ketamine analgosedation use and its effect on incident delir- adjustment and sedation–analgesia practices on severe asynchrony

ium in critically ill adults. Crit Care Explor 2021;3:e0544. DOI: in patients ventilated in assist-control mode*. Crit Care Med

10.1097/CCE.0000000000000544. 2013;41:2177-2187. DOI: 10.1097/CCM.0b013e31828c2d7a.

36. Chanques G, Constantin JM, Devlin JW, et al. Analgesia and seda-
24. Pun BT, Badenes R, Heras La Calle G, et al. Prevalence and
tion in patients with ARDS. Intensive Care Med 2020;46:2342-
risk factors for delirium in critically ill patients with COVID-
2356. DOI: 10.1007/s00134-020-06307-9.
19 (COVID-D): a multicentre cohort study. Lancet Respir Med
2021;9:239-250. DOI: 10.1016/S2213-2600(20)30552-X. 37. Carson SS, Kress JP, Rodgers JE, et al. A randomized trial of
intermittent lorazepam versus propofol with daily interruption in
25. Balas MC, Tan A, Mion LC, et al. Factors associated with spon-
mechanically ventilated patients*. Crit Care Med 2006;34:1326-
taneous awakening trial and spontaneous breathing trial perfor-
1332. DOI: 10.1097/01.CCM.0000215513.63207.7F.
mance in adults with critical illness. Chest 2022;162:588-602.
DOI: 10.1016/j.chest.2022.01.018. 38. Pandharipande PP, Pun BT, Herr DL, et al. Effect of sedation
with dexmedetomidine versus lorazepam on acute brain dys-
26. Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safety of a
function in mechanically ventilated patients: the MENDS ran-
paired sedation and ventilator weaning protocol for mechan-
domized controlled trial. JAMA 2007;298:2644. DOI: 10.1001/
ically ventilated patients in intensive care (Awakening and
jama.298.22.2644.
Breathing Controlled trial): a randomised controlled trial. Lancet
2008;371:126-134. DOI: 10.1016/S0140-6736(08)60105-1. 39. Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine
versus midazolam for sedation of critically ill patients. JAMA
27. Aitken LM, Kydonaki K, Blackwood B, et al. Inconsistent relation-
2009;301:489-499. DOI: 10.1001/jama.2009.56.
ship between depth of sedation and intensive care outcome: sys-
tematic review and meta-analysis. Thorax 2021;76:1089-1098. 40. Jakob SM. Dexmedetomidine versus midazolam or propo-
DOI: 10.1136/thoraxjnl-2020-216098. fol for sedation during prolonged mechanical ventilation: two

NEJM EVIDENCE 10
 randomized controlled trials. JAMA 2012;307:1151. DOI: trial. Lancet 2016;388:1377-1388. DOI: 10.1016/S0140-
10.1001/jama.2012.304. 6736(16)31637-3.

41. Hughes CG, Mailloux PT, Devlin JW, et al. Dexmedetomidine 54. Bailey P, Thomsen GE, Spuhler VJ, et al. Early activity is feasible and
or propofol for sedation in mechanically ventilated adults with safe in respiratory failure patients*. Crit Care Med 2007;35:139-
sepsis. N Engl J Med 2021;384:1424-1436. DOI: 10.1056/NEJ- 145. DOI: 10.1097/01.CCM.0000251130.69568.87.
Moa2024922.
55. Pohlman MC, Schweickert WD, Pohlman AS, et al. Feasibility of
42. The Sedation Practice in Intensive Care Evaluation (SPICE) Study physical and occupational therapy beginning from initiation of
Group investigators, Shehabi Y, Chan L, et al. Sedation depth and mechanical ventilation*. Crit Care Med 2010;38:2089-2094. DOI:
long-term mortality in mechanically ventilated critically ill adults: 10.1097/CCM.0b013e3181f270c3.
a prospective longitudinal multicentre cohort study. Intensive Care
56. Bourdin G, Barbier J, Burle JF, et al. The feasibility of early physical
Med 2013;39:910-918. DOI: 10.1007/s00134-013-2830-2.
activity in intensive care unit patients: a prospective observational
43. Casamento A, Neto AS, Lawrence M, et al. Delirium in ventilated one-center study. Respir CARE 2010;55:400-407.
patients receiving fentanyl and morphine for Analgosedation: find-
57. The TEAM Study Investigators and the ANZICS Clinical Trials
ings from the ANALGESIC trial. J Crit Care 2023;77:154343. DOI:
Group. Early active mobilization during mechanical ventilation in
10.1016/j.jcrc.2023.154343.
the ICU. N Engl J Med 2022;387:1747-1758. DOI: 10.1056/NEJ-
44. Ritchie BM, Torbic H, DeGrado JR, Reardon DP. Sedation vari- Moa2209083.
ability increases incidence of delirium in adult medical intensive
care unit patients at a tertiary academic medical center. Am J Ther 58. Kho ME, Berney S, Pastva AM, et al. Early in-bed cycle ergometry

2019;26:e92-e95. DOI: 10.1097/MJT.0000000000000455. in mechanically ventilated patients. NEJM Evid 2024;3(7). DOI:
10.1056/EVIDoa2400137.
45. Bernard-Valnet R, Favre E, Bernini A, et al. Delirium in adults with
COVID-19–related acute respiratory distress syndrome: compari- 59. Paton M, Chan S, Tipping CJ, et al. The effect of mobilization at
son with other etiologies. Neurology 2022;99:e2326-e2335. DOI: 6 months after critical illness — meta-analysis. NEJM Evid 2023;2(2).
10.1212/WNL.0000000000201162. DOI: 10.1056/EVIDoa2200234.

46. Andersen-Ranberg NC, Poulsen LM, Perner A, et al. Haloperi- 60. Parry SM, Nydahl P, Needham DM. Implementing early physical
dol for the treatment of delirium in ICU patients. N Engl J Med rehabilitation and mobilisation in the ICU: institutional, clinician,
2022;387:2425-2435. DOI: 10.1056/NEJMoa2211868. and patient considerations. Intensive Care Med 2018;44:470-473.
DOI: 10.1007/s00134-017-4908-8.
47. Girard TD, Exline MC, Carson SS, et al. Haloperidol and ziprasi-
done for treatment of delirium in critical illness. N Engl J Med 61. Dirkes SM, Kozlowski C. Early mobility in the intensive care
2018;379:2506-2516. DOI: 10.1056/NEJMoa1808217. unit: evidence, barriers, and future directions. Crit Care Nurse
2019;39:33-42. DOI: 10.4037/ccn2019654.
48. Su X, Meng ZT, Wu XH, et al. Dexmedetomidine for preven-
tion of delirium in elderly patients after non-cardiac surgery: 62. Kamdar BB, Combs MP, Colantuoni E, et al. The association of
a randomised, double-blind, placebo-controlled trial. Lancet sleep quality, delirium, and sedation status with daily participa-
2016;388:1893-1902. DOI: 10.1016/S0140-6736(16)30580-3. tion in physical therapy in the ICU. Crit Care 2016;20:261. DOI:
10.1186/s13054-016-1433-z.
49. Qu JZ, Mueller A, McKay TB, et al. Nighttime dexmedetomidine
for delirium prevention in non-mechanically ventilated patients 63. Thomsen GE, Snow GL, Rodriguez L, Hopkins RO. Patients with
after cardiac surgery (MINDDS): a single-centre, parallel-arm, respiratory failure increase ambulation after transfer to an inten-
randomised, placebo-controlled superiority trial. eClinicalMedi- sive care unit where early activity is a priority. Crit Care Med
cine 2023;56:101796. DOI: 10.1016/j.eclinm.2022.101796. 2008;36:1119-1124. DOI: 10.1097/CCM.0b013e318168f986.

50. Morris PE, Griffin L, Thompson C, et al. Receiving early mobility 64. Alexandre L, Michael D, Bruno M, et al. A communication strategy
during an intensive care unit admission is a predictor of improved and brochure for relatives of patients dying in the ICU. N Engl J
outcomes in acute respiratory failure. Am J Med Sci 2011;341:373- Med 2007;356:469-478. DOI: 10.1056/NEJMoa063446.
377. DOI: 10.1097/MAJ.0b013e31820ab4f6.
65. White DB, Angus DC, Shields AM, et al. A randomized trial of a
51. Burtin C, Clerckx B, Robbeets C, et al. Early exercise in critically ill family-support intervention in intensive care units. N Engl J Med
patients enhances short-term functional recovery*. Crit Care Med 2018;378:2365-2375. DOI: 10.1056/NEJMoa1802637.
2009;37:2499-2505. DOI: 10.1097/CCM.0b013e3181a38937.
66. Jabre P, Belpomme V, Azoulay E, et al. Family presence during car-
52. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physical diopulmonary resuscitation. N Engl J Med 2013;368:1008-1018.
and occupational therapy in mechanically ventilated, critically ill DOI: 10.1056/NEJMoa1203366.
patients: a randomised controlled trial. Lancet 2009;373:1874-
67. Rosa RG, Falavigna M, Da Silva DB, et al. Effect of flexible family
1882. DOI: 10.1016/S0140-6736(09)60658-9.
visitation on delirium among patients in the intensive care unit: the
53. Schaller SJ, Anstey M, Blobner M, et al. Early, goal-directed mobil- ICU visits randomized clinical trial. JAMA 2019;322:216. DOI:
isation in the surgical intensive care unit: a randomised controlled 10.1001/jama.2019.8766.

NEJM EVIDENCE 11
68. Prime D, Arkless P, Fine J, Winter S, Wakefield DB, Scatena R. 71. Mohand-Saïd S, Lalonde MR, Boitor M, Gélinas C. Family mem-
Patient experiences during awake mechanical ventilation. J Com- bers’ experiences with observing pain behaviors using the criti-
munity Hosp Intern Med Perspect 2016;6:30426. DOI: 10.3402/ cal-care pain observation tool. Pain Manag Nurs 2019;20:455-461.
jchimp.v6.30426. DOI: 10.1016/j.pmn.2018.11.001.

69. Treggiari MM, Romand JA, Yanez ND, et al. Randomized trial 72. Shahid A, Sept BG, Longmore S, et al. Development and preclinical
of light versus deep sedation on mental health after critical testing of the critical care pain observation tool for family caregiver
illness*. Crit Care Med 2009;37:2527-2534. DOI: 10.1097/ use (CPOT-Fam). Health Sci Rep 2023;6:e986. DOI: 10.1002/
CCM.0b013e3181a5689f. hsr2.986.

70. Puntillo KA, Neuhaus J, Arai S, et al. Challenge of assessing symp- 73. Putowski Z, Rachfalska N, Majewska K, Megger K, Krzych Ł. Identifi-
toms in seriously ill intensive care unit patients: can proxy report- cation of risk factors for post-intensive care syndrome in family mem-
ers help?*. Crit Care Med 2012;40:2760-2767. DOI: 10.1097/ bers (PICS-F) among adult patients: a systematic review. Anaesthesiol
CCM.0b013e31825b94d8. Intensive Ther 2023;55:168-178. DOI: 10.5114/ait.2023.130831.

NEJM EVIDENCE 12