Analytic Review

Journal of Intensive Care Medicine

2024, Vol. 39(8) 715-732
Easing Suffering for ICU Patients and Their © The Author(s) 2023
Article reuse guidelines:
sagepub.com/journals-permissions
Families: Evidence and Opportunities for DOI: 10.1177/08850666231204305
journals.sagepub.com/home/jic
Primary and Specialty Palliative Care in
the ICU

Christine Doherty, MD, MS1,2, Shelli Feder, PhD, APRN3,

Sarah Gillespie-Heyman, APRN4, and Kathleen M. Akgün, MD, MS2,5

Abstract
Intensive care unit (ICU) admissions are often accompanied by many physical and existential pressure points that can be extraor-
dinarily wearing on patients and their families and surrogate decision makers (SDMs). Multidisciplinary palliative support, including
physicians, advanced practice nurses, nutritionists, chaplains and other team members, may alleviate many of these sources of
potential suffering. However, the palliative needs of ICU patients undoubtedly exceed the bandwidth of current consultative spe-
cialty palliative medicine teams. Informed by standard-of-care palliative medicine domains, we review common ICU symptoms
(pain, dyspnea and thirst) and their prevalence, sources and their treatment. We then identify palliative needs and impacts in
the domains of communication, SDM support and transitions of care for patients and their families through their journey in
the ICU, from discharge and recovery at home to chronic critical illness, post-ICU disability or death. Finally, we examine the
evidence for strategies to incorporate specialty palliative medicine and palliative principles into ICU care for the improvement
of patient- and family-centered care. While randomized controlled studies have failed to demonstrate measurable improvement
in pre-determined outcomes for patient- and family-relevant outcomes, embracing the principles of palliative medicine and assur-
ing their delivery in the ICU is likely to translate to overall improvement in humanistic, person-centered care that supports
patients and their SDMs during and following critical illness.

Keywords
Palliative care, intensive care unit (ICU), symptoms, communication, surrogate decision maker, goal-concordant care

Introduction support patients’ and families’ distress during acute and

chronic critical illness, as well as at the end of life.
Intensive care unit (ICU) admissions during the final month of Multidisciplinary palliative care is care centered on alleviat-
life are becoming increasingly common.1,2 Approximately one ing suffering among patients with serious illness and can be
in five individuals in the United States (U.S.) will die following delivered by the primary ICU team or by consultant palliative
an ICU admission.3 Although ICU mortality is improving, many
survivors are discharged to subacute or long-term care settings
rather than home, especially among older populations.4,5 Even 1
Department of Internal Medicine New Haven, Yale New Haven Hospital, New
among survivors, short-, intermediate- and long-term outcomes Haven, CT, USA
2
such as hospital readmission and mortality remain discourag- Yale School of Medicine, New Haven, CT, USA
3
Yale University School of Nursing, Orange, CT, USA
ingly high, especially among medical ICU survivors.6-8 4
Hospice and Palliative Medicine, VA-Connecticut Healthcare System, West
Compared with non-ICU hospital survivors, 5-year readmission Haven, CT, USA
was 65% among ICU survivors (vs 49.7% non-ICU) with the 5
Section of Pulmonary, Critical Care and Sleep Medicine, Department of
majority of readmissions occurring in the first year after dis- Medicine, VA-Connecticut and Yale University School of Medicine, New Haven,
charge.8 Similarly, 5-year mortality was 32% among ICU survi- CT, USA
vors (vs 22% non-ICU).8 Moreover, these patients often Received June 25, 2023. Received revised August 30, 2023. Accepted
experience significant functional impairments months-to-years September 13, 2023.
after discharge.9-12 Therefore, it is imperative that ICU clinicians
Corresponding Author:
prepare patients and families for the issues associated with pro- Kathleen M. Akgün, MD, MS, Section of Pulmonary, Critical Care and Sleep
longed recovery or lack of recovery altogether. This requires Medicine, 950 Campbell Ave, West Haven, CT, USA.
ICU clinicians develop communication skills to anticipate and Email: [email protected]
716 Journal of Intensive Care Medicine 39(8)

medicine services13 and may provide valuable support for Defining the Scope of the Problem—What Are Potential
patients and families. ICU palliative care encompasses five Palliative Care Needs in the ICU?
core domains—symptom management, communication, surro-
gate decision maker (SDM) support, transitions of care, and While delivering LSTs to address life-threatening medical con-
workforce support [Table 1; Figure 1].13,14 For the purposes ditions, symptom burden can be quite high due to the critical
of this review, we will focus on the first four domains. (Of illness itself; discomfort also often accompanies necessary
note, the term SDM is used throughout this review to reference ICU procedures. Symptoms experienced by ICU patients are
the social supports that matter most to the ICU patient when complex and multidimensional. A full explication of all possi-
making serious medical decisions and does not intend to ble symptoms is beyond the scope of this review. We will
ascribe to heteronormative legal relationships as the standard focus on the three most common distressing symptoms reported
for SDMs). A common misconception is that specialist pallia- by ICU patients: pain, dyspnea, and thirst (Table 2).18,19 We
tive care involvement is only necessary when life-prolonging will provide an overview of symptom prevalence and manage-
treatments have “failed” and death is inevitable. In fact, pallia- ment, as well as their impact on palliative-related principles in
tive care can and should be provided concurrently with life the ICU.
sustaining treatments (LSTs), a sentiment supported by profes-
sional organizations including the American Thoracic Society Pain. Pain is common in the ICU from critical illness and the
and the American College of Critical Care Medicine.15-17 delivery of LSTs to attempt to treat critical illness, especially
This article will first review the evidence regarding the unmet among those on mechanical ventilation.20 Pain prevalence
palliative care needs of ICU patients and their SDMs primarily and severity may vary with specific patient populations. Early
framing them in relation to established ICU palliative domains. studies exploring surgical ICU survivors’ experiences of pain
We will then discuss the evidence for palliative interventions found that 7021 −77% of patients recalled uncontrolled
in the ICU that have been employed to meet these needs and pain.22 Beyond surgical ICU patients, larger scale studies
propose future strategies to deliver high-quality palliative care have found a lower, although still substantial, prevalence of
in the ICU. pain in other ICU patients. Among 171 patients at high risk
of dying (defined by severity of illness (APACHE II) score,
ICU length of stay and chronic conditions) with 405
Table 1. Five Core Domains of ICU Palliative Care. symptom assessments from two ICUs in a tertiary medical
Domain Description center in the U.S., pain prevalence was estimated at 40%.23
Among 50 cancer patients in the ICU who were able to self-
Symptom Management Managing physical, psychological, and report, 75% reported any discomfort, and 56% described mod-
spiritual distress in patients erate to severe pain, especially with ICU-related procedures.19
Communication Discussing goals of care with patients and Sources of pain range from the seemingly mundane, usual
their surrogates to ensure treatment is
in line with their values, priorities, and
ICU daily care (endotracheal suctioning and turning in the
wishes bed) to similarly routine blood tests and procedures (drawing
Surrogate Decision Supporting patient’s loved ones/SDMs arterial blood gases or placing central or peripheral venous cath-
Maker Support eters). A study in France of 1381 patients across 44 ICUs eval-
Transitions of Care Coordinating transitions of care across uated pain scores prior to and during endotracheal suctioning or
various settings (in-patient hospice, mobilization. Significantly more patients reported pain during
home palliative care etc) versus prior to delivery of these typical ICU care processes.
Workforce Support Support healthcare workers as they
navigate grief and distress
In another study of more than 6000 patients from 167 ICUs
across the U.S., United Kingdom (UK), Canada, and

Table 2. Common ICU Symptom Overview of Prevalence, Treatments and ICU Pearls.

Symptom Estimated Prevalence Treatment Notes

Pain 40–77%20-23 Opioids − Bowel regimen to prevent opioid-related constipation

Dyspnea 33–44% 19,23,32,33 Repositioning − Lower doses of opioids required to relieve dyspnea compared with pain
Fan blowing − Bowel regimen to accompany opioid use
Cold washcloth − Benzodiazepines sparingly in more severe cases37,38
Oxygen, NIV, intubation
Opioids18
Thirst 30–70%23,39,40 Mouth swabs − Untreated, contributes to anxiety and desperation for patients
Artificial saliva
“Thirst bundle”43
Doherty et al. 717

pre-procedure pain-targeting medication treatment is uncom-

mon, only occurring in 17–25% of patients262.4
Opioids are the mainstay of pain treatment in the ICU,
including for pre-procedure prevention of discomfort and pain
(Table 2).20,27 Understandably, there are often concerns about
the potential side effects of analgesic medications in critically
ill patients such as vasoplegia, respiratory depression, aspira-
tion, and constipation/ileus. However, in addition to the acute
psychological benefits to recognizing and treating critically ill
patients’ unmanaged physical pain, reducing their physiological
stress can potentially lessen the harms of their primary ICU con-
dition. Safe use of opioids requires routine monitoring of bowel
movements and their frequency; prescribing bowel regimens is
essential to prevent constipation and ileus, although evidence of
benefit in the critical care setting has not been conclusively
demonstrated.28,29
Among physical symptoms assessed, standardized scales to
Figure 1. Five core domains of ICU palliative care. monitor and treat pain may be most familiar for ICU clinicians.
Using standardized measurements for symptom assessment
Table 3. Procedure Related Pain. may translate into meaningful improvements in patients’ care.
One study examined outcomes among 230 ICU patients
Approximate procedure Average before and after implementation of a standardized pain and agi-
frequency during an pain intensity tation scoring system—with predefined thresholds for nurses to
Procedure ICU admission (out of 10) notify physicians of the patient’s pain or agitation score. The
Chest Tube Removal Occasionally 5 baseline phase included twice/day patient assessments for
Wound Drain Occasionally 4.5 pain and agitation by six independent observers (medical and
Removal pharmacy students). During the intervention, clinical staff (phy-
Arterial Line Frequent 4 sicians, residents, and nurses) were trained on vigorously
Insertion assessing and treating pain based on patient-reported pain
Suctioning Routine among ventilated 4
patients
scores; didactic sessions were also part of the unit education
Turning Routine 3 addressing risks and benefits of ICU pain management.
Positioning Routine 3 Following the intervention, pain was significantly lower (42%
Blood draws Routine 3 vs baseline = 63%). There was also a marked decrease in days
Respiratory Routine 3 of mechanical ventilation and nosocomial infection rates com-
exercises pared to the baseline group (Table 4). These differences may
Wound care Routine 3 be due to a decrease in the stress response associated with
Mobilization Routine 2
pain or more rapid de-escalating of unnecessary treatments in
Adapted from K.A. Puntillo et al25 the absence of pain or agitation.30 Another study examining
tailored approaches to relieve discomfort for ICU patients in
areas including noise, light, pain, isolation and communication
Australia, pain before and during routine ICU care (including needs (n = 16 total items) demonstrated significant improve-
general care, tracheal suctioning, turning, femoral sheath ments in all items addressed except pain.31 This may be under-
removal, and central line placement) was examined. Average scoring the multidimensional aspects of pain assessment and
pain scores ranged from 26 to 49 (on a 100-point scale; management in the ICU that require more specific, targeted
higher numbers indicate worse pain).24 Pain scores were signif- interventions. Regardless, these results illustrate the feasibility
icantly higher across all procedures relative to baseline; turning and effectiveness of improving patient-centered symptom
had the highest reported pain intensity (pain score = 49/100) management in the ICU with protocolized assessments, man-
and femoral sheath removal had the lowest (pain score = 26/ agement and education by palliative medicine-minded unit
100). A more recent study of 3851 ICU patients from 192 champions.20,30
ICUs worldwide examined pain intensity during several proce-
dures (Table 3).25 The three most painful procedures were chest Dyspnea. Dyspnea remains a common, uncomfortable, underre-
tube removal, wound drain removal and arterial line insertion. cognized and undertreated symptom in the ICU. Between 33
Although average baseline pain was mild for patients (1-4 on and 44% of ICU patients report dyspnea which often co-exists
a 0-10 scale, higher scores conferring greater intensity), all pro- with anxiety.19,23,32,33 ICU survivors report dyspnea as one of
cedures caused a significant increase in pain compared to base- the worst ICU-related memories.34 Despite its distressing
line.25 Despite the frequency of procedure-related pain, nature, high prevalence and contributions to additional
718
Table 4. Key Studies on Palliative Interventions in the ICU and Associated Outcomes and Implications.

Study/citation Years Country(ies) and Patient population Palliative intervention Outcomes associated with intervention(s)

Symptom management
Chanques 200630 11/2002–3/2003 France (single medical-surgical ICU) Behavior pain scale, numeric Decreased pain (63% vs 42%; p = .002)
4/2003–10/2003 Control n = 100 rating scale, Richmond Decreased severe pain (36% vs 16%, p < .001)
Intervention n = 130 agitation sedation scale per Decreased severe agitation)18 versus 5%, p = .002)
nursing shift with physician Intervention associated with:
notification Decreased ventilator days
Decreased infections
More escalation and de-escalation of medications
Greater nurse satisfaction
Kalfon 201731 10/2014–10/2015 France (34 ICUs; medical, surgical, mixed) Tailored multi-component Reduced patient-reported discomfort score from
Patients 18 years and older who survived 3 program addressing 16-item IPREA questionnaire by 6.35 points, SE
calendar days discomfort in 16 areas 1.23, p < .001 in adjusted analyses
Unexposed n = 360 measured on day of ICU All items reduced except pain
Exposed n = 398 discharge; immediate Greatest impact on thirst and lack of intimacy
feedback to bedside nurse,
electornic reminders of
findings, monthly unit
reports and themes
Program led by
physician-nursing ICU duo in
exposed units
Kalfon 201995 10/2014–10/2015 France (Extension of IPREA3 study) Tailored multi-component PTSD assessed by Impact of Event Scale (IES)-Revised
1537 ICU survivors from 30 ICUs; program for at least 5 PTSD lower at one-year post-discharge (6.1% vs
1-year outcome data: Unexposed n = 475 months (from Kalfon 2017 12.0%; in adjusted analyses p-.15)
Exposed n = 344 study) Models with IER-R imputations with higher scores in
unexposed group (pooled OR = 1.55 [1.03-2.33))

Communication
Lautrette 2007122 5/2005–10/2005 France (22 ICUs) Proactive family conferences Decreased 90-day family-reported symptoms of
Patients whose attending physician believed and bereavement brochures − PTSD (45% vs 69%; p = .01); lower IES scores
patient would die in coming days (27 vs 39; p = .02)
Control n = 52 − Depression (29% vs 56%; p = .003)
Intervention n = 56 − Anxiety (45% vs 67%; p = .02)
Curtis 2008117 7/2003–3/2004 USA (single 350-bed hospital) Clinician education and training, No difference in family-assessed quality of dying
4/2004–11/2004; 12/ Patients who died in ICU with minimum 6 h ICU feedback to physicians, staff though trend towards improved (67.1 vs 62.3;
2004–10/2005 stay or within 24 h transferred elsewhere in support, system support, p = .09)
hospital “Get to know me” posters, Improved nurse-assessed quality of dying
Pre-intervention n = 253 family pamphlets (67.1 vs 63.1; p < .01)
Post-intervention n = 337 Decreased ICU length of stay (3.1 vs 3.9; p < .01)

(continued)
 Table 4. (continued)

Study/citation Years Country(ies) and Patient population Palliative intervention Outcomes associated with intervention(s)
118
Curtis 2011 5/2004–2/2008 USA Clinician/hospital-level teachingNo difference in family- or nurse-assessed quality of
12 hospitals (6 controls, 6 intervention) using self-efficacy theory to dying
− 1239 baseline patients promote ICU palliative care No effect on ICU length of stay or time from ICU
− 1079 follow up patients (see Curtis 2008 admission to withdrawal of LSTs
− 43% family response intervention) Decreased family meetings, less prognosis discussion
Curtis 2013119 2007–2013 USA (Univ. Washington and MUSC) 8-session simulation-based 1866 patient ratings, 936 family ratings
391 internal medicine and 81 nurse practitioner communication skills No improvement in quality of communication or
learners quality of end of life communication per patients or
families
More depression for patients from postintervention
trainees (PHQ-8 10 vs 8.8 pre-intervention;
adjusted intervention effect 2.2, p = .006)
Curtis 2016120 11/2008–10/2013 USA (2 hospitals) Facilitators supporting Decreased depressive symptoms at 6-months
Patients with predicted mortality ≥30% and with communication between (p = .017); No effect at 3 months
SDM clinicians and SDMs/families No effect at any time point for anxiety or PTSD
No difference in mortality
Decreased ICU costs, especially among decedents
Among decedents: intervention decreased ICU
(7.7 vs 28.5) and hospital (8.0 vs 31.8) length of stay
Carson 201679 2010–2014 USA (4 MICUs) Palliative specialist-led No difference in anxiety or depression
Patients ≥21 years with 7 days of mechanical structured family meetings More PTSD symptoms in intervention (IES-R 25.9 vs
ventilation (at least 2) and informational control = 21.3; p = .0495)
Control n = 126 with 184 SDMs brochure No difference in discussion of patient preferences
Intervention n = 130 with 184 SDMs No difference in hospital length of stay or 90-day
survival
White 2018121 7/2012–2/2016 USA (5 ICUs in single health system) Multicomponent family-support No difference in SDM anxiety or depression scores
1420 eligible ICU patients (1106 agreed to intervention delivered by No difference in SDM impact of event scores
long-term follow up, with 73% SDMs contacted) ICU team Improved quality of communication and patient
age ≥18 without decision making capacity; at perception of patient centeredness
least 4 days on vent or estimated 40% chance of Shorter ICU length of stay (6.7 vs 7.4 days; p = 0045),
dying of having great disability especially decedents (4.4 vs 6.8 days, p < .001)
Chang 202162 6/2017–12/2019 USA (3 California hospitals) Communication training with More frequent discussions of risks/benefits of ICU
Preintervention n = 113 time-limited trial default treatment (94.9% vs 34.9%), more eliciting values/
Post-intervention n = 96 framing and care planning for preferences (98.3% vs 46.5%) and clinical markers
family meetings with SDMs of improvement (88% vs 20.9%); all p < .01
Decreased ICU length of stay (7.4 vs 8.7 days, p = .02)
Decreased use of mechanical ventilation (72.9% vs
85.8%, p = .02)
Mehta 2023112 1/2019–6/2019 Single center MICU Embedded palliative care team, Early palliative evaluation
9/2019–2/2020 Retrospective evaluation daily rounds with ICU team More likely SDM documentation and goals of care
Pre n = 52 patients (traditional referral-based Less likely to die in hospital
consult)
After program implementation n = 169
Ma 2019111 8/2017–5/2018 Patients randomized to intervention versus Trigger for palliative care Trigger for palliative consult associated with
control consult − 50% DNR code status versus 23%
− 18.6% hospice versus 4.9%

719
(continued)
720
Table 4. (continued)

Study/citation Years Country(ies) and Patient population Palliative intervention Outcomes associated with intervention(s)

Intervention—97 − fewer vent days (4 vs 6)

Usual care—102 − tracheotomy 1.0% versus 7.8%
Diaries − fewer ED visits/readmissions (17% vs 39%)
Jones 201093 2006–2008 Europe (12 ICUs) Prospective diary entries by Decreased new onset PTSD for intervention group at
ICU stay >72 h ICU patients, reviewed 1 3-month follow up (5% vs 13.1%, p = .02)
Controls n = 175 month after ICU discharge Post hoc analysis greater improvement in PTSD
Intervention n = 177 symptoms from 1 to three months for those with
PTSS-14 > 45
Garrouste-Orgeas 10/2015–1/2017 France (35 ICUs) ICU diaries completed by No difference at 3-month PTSD or IES between
201994 709 patients with mechanical ventilation within clinicians and family members groups
48 h for at least 2 days
Control n = 354
Intervention n = 355
Support for SDMs of dying patients

Kentish-Barnes 12/2014–10/2015 France (22 ICUs) Condolence letter receipt from 1-month: no significant difference in depression
2017 123 242 (controls n = 119, intervention n = 123) physician/nurse in charge at 6-month: intervention associated with increased:
SDMs/relatives of patients who died the time of death − HADS (13 vs 10, p = .04)
− HADS-depressions subscale (6 vs 3, p = .02)
− depressive symptoms (36.6% vs 24.7%, p = .05)
− PTSD-related symptoms (52.4 vs 37.1, p = .03)
Kentish-Barnes 2/2017–10/2019 France (34 ICUs) 3-step support strategies Prolonged grief symptoms reduced in intervention
2022124 Controls n = 391 relatives supporting families through group 15% versus 21%, p-.035; lower median
Intervention n = 484 relatives dying process (preparation prolonged grief-13 scores
Patients 18 and older with ICU LOS 2 or more for death, during dying
days process, after death)
Abbreviations: HADS = Hospital Anxiety and Depression Scale; ICU = Intensive Care Unit; LOS = Length of Stay; OR = Odds Ratio; PHQ = Patient Health Questionnaire; PTSD = Post Traumatic Stress Disorder.
Doherty et al. 721

symptoms like anxiety, dyspnea has received considerably less decrease thirst intensity and distress in patients and can be rel-
clinical and research attention than pain.35 Even when dyspnea atively easily implemented.43
is identified by ICU teams, symptom treatment is not guaran-
teed. One observational study of 138 ICU patients found that Proxy-Reported Symptom Assessment. Although self-report is the
identification of moderate-to-severe dyspnea by clinicians was reference standard for assessing symptoms, it is not uncommon
not significantly associated with any subsequent clinical or that ICU patients are unable to communicate their symptoms
therapeutic interventions compared to identification of and needs due to their critical illness. ICU clinicians and
moderate-to-severe pain.36 patients’ families, friends and SDMs may be asked to make
Uncontrolled dyspnea during ICU admissions has lasting proxy symptom assessments. Evidence from the palliative med-
effects on patients. A recent study examined the prevalence icine and ICU literature suggests that proxy assessments corre-
of dyspnea and its potential association with post-traumatic late relatively well with self-reported symptoms.44 One small
stress disorder (PTSD) among intubated and mechanically ven- study of 31 ICU patients found moderate-to-strong correlation
tilated patients in 10 ICUs in France.33 Among 612 patients between patient and nursing pain assessments.45 Another
enrolled in the study (mechanically ventilated for >24 h and study of 171 ICU patients compared patient versus proxy
able to self-report dyspnea), 34% reported dyspnea (median reports (family, nurse, physician) on 10 common symptoms—
rating = 5 (scale 0-10, no discomfort to worse imaginable dis- pain, tiredness, dyspnea, restlessness, anxiety, sadness,
comfort, respectively). Similar to previous studies, anxiety hunger, fear, thirst and confusion. They found a high degree
prevalence and intensity were higher among patients with of agreement between patient and family member reports.
dyspnea. Among 153 patients who completed 90-day post- Although proxy reports are not the ideal, they can be a useful
discharge interviews, greater dyspnea density (dyspneic epi- tool, especially for patients who are unable to report their symp-
sodes/days from enrollment to extubation) was associated toms for themselves.46
with 7% increased odds of PTSD-related symptoms (odds Unmet symptom management for dying patients may be
ratio = 1.07 [95% confidence interval 1.01-1.13]). There were another source of suffering for patients and SDMs.
no differences in mechanical ventilation duration, hospital Undertreated symptoms at the end of life also contribute to
length of stay, and ICU, in-hospital, and 90-day mortality moral distress of healthcare clinicians. From one foundational
between patients who did or did not experience dyspnea.33 study of bereaved SDMs interviewed after their loved one’s
Dyspnea can be treated non-pharmacologically and pharma- death, 18% of SDMs indicated that the patient needed more
cologically.18,37 Simple, potentially low-risk strategies such as care or treatments to alleviate pain or other symptoms at the
repositioning in the bed or chair, blowing a fan towards the end of their life.47 In another study that interviewed families
patient or placing a cold washcloth on their face can relieve and caregivers following their loved ones’ death (36% in hospi-
dyspnea for some patients (Table 2). Oxygen, noninvasive ven- tal of 1578 deceased patients enrolled), inadequate or over-
tilation and mechanical ventilation may also play a role in looked pain or dyspnea management was reported for 19.3%
relieving dyspnea or bridging symptoms while additional and 18.9% of participants, respectively.48 Another study inter-
goals of care decisions are addressed. Depending on the viewing ICU nurses about the deaths of 200 ICU patients
patient’s clinical condition, bronchodilator medications and reported that almost half of the patients suffered severely in
diuretics could reduce dyspnea symptoms. In the absence of their final week of life due to symptoms including pain,
relief with these approaches, opioids are the primary pharmaco- dyspnea, edema, loss of dignity, and loss of mobility.49
logic treatment of dyspnea. Importantly, compared with pain Despite the high prevalence and symptom burden of the ICU
relief, lower doses of opioids are required for dyspnea relief. experience, there is currently no standardized, comprehensive
Benzodiazepines may play a role in patients who are also suf- approach across ICUs for assessing and treating a wide range
fering from anxiety but should be used sparingly and with of symptoms. Protocolized proxy symptom assessment and
great caution in conjunction with opioids.37,38 management for dying patients remain areas for improving
patient-centered ICU care. Integrating palliative care principles
Thirst. Thirst has been reported by 30–70% of ICU patients but and consultants into ICU care creates the opportunity to
may be similarly underrecognized by bedside nurses and physi- improve routine, daily assessment and management of patients’
cians.23,39,40 Among ICU cancer patients, 71% self-reported uncontrolled symptoms. Focusing on improved symptom
moderate or severe unsatisfied thirst (Table 2).19 Among assessment and management are achievable targets to relieve
patients with chronic critical illness, thirst prevalence was patient suffering in the ICU.
even higher, with 32/36 patients, nearly 90%, endorsing unsat-
isfied thirst.41 Thirst is associated with primary critical care con- Quality Communication—An Unmet Opportunity. Another key
dition, mechanical ventilation, oxygen mask use and a variety tenet of palliative care is facilitating quality communication
of medications. Soberingly, qualitative interviews of ICU survi- between the patient, SDM, and the primary medical team.
vors revealed that unsatisfied thirst contributed to feelings of Communication is particularly important in the ICU where
desperation, anxiety and loss of control and powerlessness.42 patients and their SDMs are often required to make difficult
A “thirst bundle” in ICU patients including sprays/swabs of life or death decisions urgently or emergently. In a recent
cold sterile water and the use of mouth/lip moisturizer may scoping review of priorities in older persons towards end of
722 Journal of Intensive Care Medicine 39(8)

life (EOL), patients identified effective communication as a key review. Investigators found that the degree of concordance
component to the quality of their care.50 depended on the specific treatment. For resuscitation, the con-
Quality communication can be difficult to assess or cordance rate was 98%; other medical interventions and antibi-
quantify. The ICU environment may be more susceptible to otics were also > 90%. However, the concordance rate for
miscommunications or insensitive communication between feeding tubes was only 64%. Although the POLST form
the treatment teams and SDMs. Among bereaved family allows for a “defined trial period of feeding tubes”, all of the
members, 23% could not recall discussions of treatment deci- patients with discrepant orders had feeding tubes in place for
sions at the EOL; better communication was one of the main > 30 days.58,59 In a separate retrospective study of 198 patients
recommendations for improving EOL care of their loved comparing preferences documented on registered advanced
ones.47,51 While it is possible dying was not discussed at all, care planning forms and hospital care over a 1-year period, con-
it is also possible that lack of recall for these conversations cordance between stated preferences in advanced care plans and
could be explained by ICU teams failing to be explicit. White hospital care was 78% with surgical procedures, IV fluids, and
et al. audio recorded 51 physician-family conferences where antibiotics accounting for most discordant care.60
EOL treatment was discussed; more than one-third of the An important caveat to interpreting the evidence base of ICU
time, physicians did not directly discuss the patient’s prognosis goal-concordant care delivery is the very nature of medical
for survival.52 Even when difficult topics like dying are dis- decision making, where goals and preferences change over
cussed between physicians and family members, there are sig- time, often as the scope of treatment options narrows.
nificant discrepancies between what clinicians think they Determining whether patients receive goal-concordant care is
communicated and what is understood by the patient or more complicated that simply comparing written documents
SDM.53,54 Discordance between primary language spoken by of preferences at one time point to medical records of care deliv-
clinician teams and patients/SDMs may further exacerbate mis- ered at another. Often decision making can change when faced
communication and sow confusion. Culturally competent, with an acute health decompensation. Real-time decision
patient- and family-centered communication skill development making can look considerably different from previously articu-
remains a lifelong mission for ICU clinicians, regardless of pal- lated and documented preferences. These dynamic decisions are
liative specialty presence. difficult to capture with chart review alone and may alter mea-
surement of goal-concordant care, highlighting the complexity
Aligning Care with Patients or SDM Preferences and Values—An of determining goal-concordant care for research purposes.61
Unmet Goal. High-quality communication harmonizing realis- Understanding factors associated with regret for SDMs in
tic treatment options with the preferences and values of patients the ICU may begin to shine light on the impact of perceived
is essential to delivering goal-concordant ICU care. goal-concordant care (and lack thereof). Among 748 SDMs par-
Goal-concordant care is a quality metric used by the ticipating in the PARTNER 2 trial, 54% reported mild regret
American Academy of Hospice and Palliative Care Medicine and 19% reported moderate-to-strong regret 6 months after
in evaluating care provided to seriously ill patients.55 In deliv- the ICU event.56 Factors associated with regret included poor
ering goal-concordant care, clinicians aim to ensure that the patient outcomes (death or severe functional impairment) and
interventions and LSTs patients receive reflect the patients’ limitations on life support. SDM black race was also associated
values and preferences, either explicitly stated or through infer- with regret. While addressing racial disparities in ICU palliative
ences made by their SDMs. In secondary analysis of data from care is beyond the scope of this manuscript, this finding high-
the PARTNER 2 study examining the efficacy of an ICU family lights the persistent and urgent need to build a racially and eth-
support intervention, goal-concordant care was associated with nically diverse workforce to deliver optimal patient- and
decreased anxiety, depression, trauma, and regret among family-centered, goal-concordant ICU palliative care.
SDMs.56 Goal-concordant care can also improve trust with Strategies such as time-limited trials may be useful for
the medical team, satisfaction with the care being provided, helping patients/SDMs feel supported in the decision-making
and create a greater sense of peace.57 while also reassessing ICU treatment preferences as the
However, patients may not always receive goal-concordant course unfolds. Training clinicians to use time-limited trial lan-
care at the EOL. A seminal study that explored the perceptions guage and counseling has been associated with more empathic
of the dying experience through SDM interviews found that communication and reduced ICU and ventilator days
almost 60% of patients preferred comfort-focused treatment at (Table 4).62 Familiarity with time-limited trials, which inher-
the EOL. Unfortunately, 10% of patients received care contrary ently acknowledges clinical uncertainty, may promote align-
to the patient’s preferred approach.51 Moreover, patients fre- ment in decision making between patients/SDMs and ICU
quently received LSTs with 25% dying on a ventilator, 11% teams.
undergoing unsuccessful cardiopulmonary resuscitation and
40% dying with a feeding tube in their last three days of Potential Palliative Needs of Family Members/SDMs During and
life.51 More recently Hickman and colleagues evaluated the After the ICU. Critical illness impacts families as well as
consistency between Physician Orders for Life-Sustaining patients. Psychological symptoms including depression and
Treatment (POLST) and treatments provided among 870 complicated grief can occur frequently among family
nursing facility residents over a 60-day period via chart members of patients for months to years from the original
Doherty et al. 723

event.63-66 In a systematic review to determine the prevalence leading to the dependence on mechanical ventilation for more
of depression, anxiety, and PTSD symptoms in caregivers of than 6 h per day for at least 21 consecutive days (with some
ICU patients (included 39 studies, 22 with surrogate/family earlier definitions including placement of tracheotomy).70
members symptoms during ICU admission; sample sizes Using this definition, approximately 5%–10% of patients who
range: 20-809, loss to follow-up = 4-77%), the prevalence of require mechanical ventilation for any reason will develop
depression ranged from 14 to 95% and anxiety 0–73%.67 CCI.71
Twenty-nine studies evaluated caregiver symptoms after ICU Over time, the definition of PCI/CCI has been broadened to
discharge (follow-up ranging from 1 to 53 months). Again, include conditions other than prolonged mechanical ventilation.
there was significant variability with depression ranging from A newer consensus definition identifies persistent critical illness
4% to 94%, anxiety from 2% to 80%, and PTSD from 3% to as at least 8 days in the ICU with any of the following: mechan-
62%. Potential explanations for the wide variability include het- ical ventilation for ≥96 h, tracheotomy, sepsis or other severe
erogeneity of study design, different instruments used to infection, extensive wounds, stroke, or traumatic brain
measure psychological outcomes, different follow-up times, injury.72 As more consistent definitions are applied for PCI/
and variable degrees of missing data or participant attrition. CCI, intermediate and long-term survival rates remain
Despite the variability, depression, anxiety and PTSD rates between 32% and 50%.73 For the purposes of this review, we
from several studies were substantially higher than in the use PCI and CCI interchangeably to refer to critical illness
general population (estimated prevalence of depression = 7%, that fails to improve despite more than a week of treatment.
generalized anxiety disorder = 3% and PTSD = 3.5%). One-year mortality is essentially a coin toss for patients with
In addition to psychological symptoms, caregivers often CCI.74 A study examining outcomes of prolonged mechanical
experience significant changes to the structure of their lives. ventilation reported high 1-year mortality rates in patients
Many have to stop working or significantly cut back on requiring mechanical ventilation with and without a tracheot-
hours, imposing new financial burdens on families, caregivers omy (48% and 58%, respectively).75 In a prospective cohort
and SDMs.68 Caregivers also may experience new restrictions study of 200 medical, surgical, and trauma patients with CCI
in their personal and social lives, with a large portion of their (defined as requiring mechanical ventilation for >21 days)
time centered on caring for their loved one. These effects can 1-year survival was 48%.76 A more recent study of 157 patients
be further exacerbated by the lack of recognition of their suffer- (mechanically ventilated >10 days and tracheotomy placed)
ing. One family member succinctly described the experience, found a 1-year survival rate of 55%.77
CCI patients also live with a very high symptom burden,
“I feel like I need to be there for him, but he is difficult. I know he facing many of the same symptoms experienced during acute
is scared, but he doesn’t seem to realize how hard on me this is… critical illness (Table 5). One study of patients who were under-
how hard it is for our kids. I am a nurse and see this every day, but going prolonged ventilator weaning found that 44% percent
I can’t believe this is happening to me. It is just too much some-
reported pain. Alarmingly, even with moderate or severe pain
times, overwhelming. What will I do if he gets worse? We are
sinking.”69
55% of the time, CCI patients reported that they received no
opioid analgesics on a given study day,41 with difficulty com-
Integrating primary and specialty palliative care into the ICU municating a source of severe distress.41
may be an underutilized resource to help alleviate and mitigate In addition to physical symptoms, many patients living with
psychological distress and burdens for caregivers. CCI have a low reported quality of life. Gardner et al evaluated
the outcomes of sepsis patients in trauma or surgical ICUs who
Transitions of Care developed CCI (defined as >14 days in the ICU) compared to
Persistent (chronic) critical illness definitions, symptom burden
those who recovered rapidly. Compared to sepsis patients
and palliative needs. Patients who survive an acute episode of
without CCI, CCI patients had worse one-year physical func-
critical illness but continue a prolonged course requiring ICU
tioning and health-related quality of life.73
care have been characterized as living with persistent critical
Between the high mortality, CCI-related symptoms, and
illness (PCI) or “chronic critical illness” (CCI). CCI was origi-
the impact on families, palliative care specialists would
nally defined by a consensus conference as respiratory failure
seem to be a fundamental support for CCI patients and their
families.74,78 Palliative care may ensure CCI patients’ symp-
Table 5. Prevalence of Symptoms Experienced by Persons Living toms and distress are adequately managed, and that goals of
With CCI. care and prognosis are clearly addressed and readdressed
as the patient’s condition changes over time. However, a
Sign or symptom Prevalence
palliative-led family meeting intervention to address the
Severe distress from inability to communicate 90% needs of CCI patients from 4 U.S. ICUs not only failed to
Thirst 80% demonstrate improvements in SDM anxiety or depression
Psychological distress 60% but also suggested potential harms with more PTSD in the
Dyspnea 60% intervention group (Table 4).79 Future studies will be
Pain 44%
needed to explore how to best integrate palliative care into
Adapted from Nelson et al, 200441 the routine care of CCI/PCI patients.
724 Journal of Intensive Care Medicine 39(8)

Transitions of Care, Psychological Burdens and Palliative Care ICU-depression.86 However, haunting or vague memories of
Needs for Post-Intensive Care Syndrome (PICS) ICU Survivors. ICU experiences appear to be significant risk factors for
Even for patients without CCI, ICU survivors often have persis- worse psychological outcomes. Poor recall of ICU stay, mem-
tent physical and psychological symptoms. The constellation of ories of nightmares, hallucinations and fearfulness were associ-
physical, psychological, and cognitive symptoms commonly ated with greater depressive symptoms; post-ICU anxiety and
experienced in the short and long-term has been coined “post PTSD symptoms and change in physical functioning after dis-
intensive care syndrome” (PICS).80 During the initial months charge also increased risk for depression.87,88,90 Two of the
following ICU discharge, 89–97% of patients report at least aforementioned studies found that depressive symptoms were
one persistent symptom, commonly pain, sleep disturbances, correlated with worse health-related quality of life (HRQOL)
fatigue and weakness.69,81,82 One year after discharge, more in ICU survivors.86,88 Alarmingly, 23% of depressed patients
than half of ICU survivors suffer from ongoing symptoms in reported suicidal ideation following their ICU stay.86 In the
at least one of the three PICS domains (physical, psychological, one study that determined antidepressant medication use pre
and cognitive) and 20% have symptoms in two or more and post ICU admission, 24% of patients were taking antide-
domains.83 Physical symptoms can persist for years for patients pressants in the week prior to ICU treatment compared with
with PICS. In a prospective longitudinal cohort study of 109 49% within two months of ICU discharge.
ARDS survivors over a 5-year period, many reported subjective Recognition of depressive symptoms in ICU survivors has
perceived weakness despite no demonstrable weakness on grown considerably over the years. Rabiee and colleagues
exam. Exercise impairments were frequently reported and conducted a meta-analysis evaluating depression in ICU sur-
median 6-min walk distance was 76% of the distance walked vivors from studies spanning 1960–2015.91 They included
by age- and sex-matched controls. ICU survivors also reported studies with adult ICU survivors who had depressive symp-
ongoing reduced physical quality of life, with scores remaining toms assessed using a validated depression questionnaire.
one standard deviation below the mean score for age- and sex- This meta-analysis, including 4113 unique patients from 38
matched controls. Of note, these data represent findings for studies, found strikingly similar results to the previous sys-
patients with a median age at study enrollment of 45 years, tematic review with point prevalence of clinically significant
77% of whom were working prior to developing ARDS, depressive symptoms of 29%–30%.91 Age and sex again
making this level of perceived disability all the more salient showed no consistent association with depressive symptoms.
and urgent to address.11 Pre-ICU psychological comorbidities, experiencing psycho-
Qualitative studies enrich our understanding of the needs of logical distress during ICU/hospital stay, and delusional
PICS survivors. In interviews with 17 survivors, Eaton and col- memories were associated with more depressive symptoms
leagues found that all but one participant reported ongoing at follow-up. Anxiety and PTSD symptoms again were corre-
physical, psychological or cognitive symptoms. Feelings of lated with depressive symptoms. Severity of illness, hospital
loss or of being a burden and need for further social support length of stay, benzodiazepine use, duration of sedation, and
were also commonly expressed.84 Many identified their ICU sedation minimization strategies did not correlate with
experience as a life-altering event affecting many individuals’ depressive symptoms.
spiritual beliefs. A study in the UK aimed to describe the prevalence of psy-
The psychological weight of recovery among ICU survivors chopathology among ICU survivors and to evaluate for poten-
is also heavy. The physical stressors patients experience tial association with mortality. Validated depression, anxiety,
coupled with loss of autonomy and difficulty communicating and PTSD questionnaires were mailed to ICU survivors from
contribute to ongoing psychological symptoms. In a systematic 31 UK ICUs 3 and 12 months post-discharge. Among 13,155
review, Davydow et al quantified the prevalence of depression, patients (age >16 years; ICU ≥24 h; survived to discharge
its potential risk factors, and its impact on health-related quality and remained alive at 3 or 12 months), 38% (n = 4943) com-
of life among ICU survivors.85 Fourteen studies with 1621 pleted at least one survey; among respondents, the 3-month
unique patients were included in the analysis (follow-up prevalence of clinically significant anxiety, depression, and
ranging from 2 weeks to 14 months). Median point prevalence PTSD symptoms were 45%, 41%, and 22%, respectively, and
for clinically significant depressive symptoms was 28% (range were largely unchanged at 12 months. More than half of respon-
8%-57%) (vs 4%-7% within the general, non-clinical popula- dents (55%) experienced clinically significant symptoms of
tions). Similarly, the median point prevalence for depression anxiety, depression, or PTSD at either time point. Patients
diagnosed by a clinician was 33% (vs 7%-point prevalence in with depressive symptoms at either time were 50% more
the general population).86 Demographic risk factors have likely to die within the first 2 years in adjusted analyses com-
been inconsistently identified. Female sex was identified in pared to those without depressive symptoms; no such associa-
one study,87 but three other studies did not find this associa- tions were observed with clinically significant anxiety or
tion86,88,89; none of the studies found an association with age. PTSD symptoms.92 These studies increase our professional
Pre-morbid psychiatric history, proxy-reported depressed confidence in estimating depression during the first year of
mood prior to ICU, and poor baseline physical functioning ICU survival and should fortify the call to develop support
were all associated with increased risk for post and treatment for patients living with post-ICU depression.
Doherty et al. 725

Longer-term data suggest that psychological symptoms Pandharipande and colleagues determined the prevalence of
persist for ICU survivors beyond the first year. In a study of cognitive impairment and its risk factors in a large, diverse pop-
109 ARDS patients from Toronto, Canada, psychological ulation of ICU survivors.99 Adults admitted to medical or sur-
symptoms and physical and clinical outcomes were determined gical ICUs with respiratory failure, cardiogenic shock or
5 years after ICU discharge. At least one episode of physician- septic shock were included; of note, there were extensive exclu-
diagnosed depression or anxiety was identified for 51% of sion criteria such as pre-existing cognitive impairment, recent
patients between 2 and 5 years of follow-up. Additional ICU stays, delirium/cognitive function that prevented reliable
mental health challenges including psychotic episodes second- assessment, visual/hearing impairment or language barriers,
ary to PTSD and severe depression with agoraphobia were also unlikely to survive their ICU stay and those for whom
experienced by survivors, though were less frequent than follow-up would be difficult due to housing insecurity, active
depression and PTSD.11 substance use, psychosis, or home > 200 miles away.
Bringing the sources of trauma to the surface may validate Delirium assessments using Confusion Assessment
patients and decrease their intermediate and longer-term psy- Method-ICU (CAM-ICU) scores and use of sedative/analgesic
chological burdens and outcomes. Diaries completed by medications were recorded for enrolled patients during their
patients during their ICU stay and reviewed 1 month after dis- hospitalization. Cognitive functioning was assessed at 3- and
charge has been associated with decreased 3-month PTSD 12-months post-discharge using the Repeatable Battery for
(Table 4).93 However, a subsequent study of diaries for patients the Assessment of Neuropsychological (RBANS). Of the 569
on mechanical ventilation completed by clinicians and SDMs surviving patients, median RBANS scores for ICU survivors
made no difference in 3-month PTSD.94 The IPREA3 cluster were 1.5 standard deviations below age-adjusted population
randomized study tested implementation of a multicomponent, means. Alarmingly, 34% of patients had global cognition
closed-loop program led by physician-nurse pairs from partici- scores similar to the degree observed among those with moder-
pating French ICUs to decrease patient discomfort. One-year ate traumatic brain injury (TBI) at 12 months and 24% had
PTSD symptoms were significantly lower for survivors from scores similar to patients with mild Alzheimer’s disease (AD)
ICUs where the program was implemented.95 There was no (2 SD below population means). These effects were not just
effect on depression or anxiety. Together, these findings at observed in older adults–patients <49 years old without preex-
the very least suggest vigilance in listening to patients/SDMs isting comorbid conditions had similar levels of global cogni-
during and after ICU admission as well as observations of tion score impairments. Longer duration of delirium was
bedside nurses and clinicians to identify strategies to mitigate independently associated with worse global cognition scores;
post-ICU psychological harms. use of sedative/analgesic medications was not. These data
Patients may experience persistent impairments in long-term make it increasingly clear that cognitive dysfunction is a
HRQOL although trajectories may be unpredictable. In another common sequela of surviving critical illness.
UK study including general ICU survivors, patients’ relatives COVID-19 ICU survivors unsurprisingly face similar phys-
rated the patient-level quality of life pre-ICU admission using ical, psychological and cognitive challenges. A large number
the short form (SF)-36 and compared with patient-reported experience weakness due to prolonged duration of mechanical
SF-36 at 3, 6, and 12 months post discharge and again at 2.5 ventilation.100 A high prevalence of depression, anxiety, acute
and 5 years.96 Of 300 patients recruited, 202 patients completed distress disorder, PTSD, and cognitive impairment have also
3-month surveys with progressive declines in response rates been reported101-104
over time due to loss to follow-up and patient deaths. Among Primary palliative care, defined as goals of care discussions,
97 patients (32%) who completed 5-year surveys, there was a basic symptom management, and care coordination provided by
marked improvement between the pre-morbid and 5-year psy- a clinician not board-certified in palliative care, is imperative to
chological HRQOL scores, with HRQOL scores not signifi- address these unmet needs of survivors. However further work
cantly different from population means. The authors postulate is needed to entrust primary palliative care skills in clinicians
that this may be due to patients feeling as though they caring for ICU survivors with appropriate tailoring and integra-
“cheated death”. It is also possible that the SF-36 does not tion of specialty palliative care into the recovery of ICU survi-
assess relevant aspects of mental health quality of life for vors with lingering and troublesome physical, psychological
patients surviving mechanical ventilation. By contrast, at and cognitive symptoms and PICS.
all-time points, physical quality of life scores were lower than
population means.96
Cognitive impairment is common among ICU survivors and Approaches for Integrating Palliative Care into ICU Care
may serve as another layer of stress for patients and families. A Access to palliative care has significantly increased over the
small study of medical ICU survivors 6-months post-discharge past 20 years. By the end of 2014, 67% of U.S. hospitals
found that 32% (11/34) had cognitive impairment on formal with 50 or more beds had palliative care programs, compared
neuropsychological testing.97 In a prospective cohort study of to only 15% in 2001.105 Not only is the number of hospitals
older ICU survivors (age 65+), there was a two-fold greater with palliative care programs increasing, the percentage of
risk for developing dementia compared to adults >65 years patients seen by palliative care teams is also increasing. From
old who were not hospitalized.98 More recently, 2008 to 2018, the percentage of hospital admissions seen by
726 Journal of Intensive Care Medicine 39(8)

palliative care teams increased from 2.5% to 5.3%.106,107 days) and had fewer post discharge ED visits or readmissions
Within the ICU, patients receive palliative care via two (17% vs 39%).111
models of delivery: consultative and integrative. We discuss Mehta and colleagues evaluated the effect of a palliative care
each model and the evidence for these models or aspects of consultation team embedded into the MICU. This embedded
care processes into the ICU. palliative care team reviewed any unmet palliative needs of
patients through chart review and interdisciplinary team discus-
sions daily and identified patients who may need a family
Consultative Model. In the consultative model, palliative care meeting. In comparison to a retrospective cohort study evaluat-
services are provided by a specialist, board-certified palliative ing outcomes before implementation of this team (n = 52),
care clinician. The consult is usually either requested by the patients who received an embedded consult (n = 169) were
ICU team or generated from pre-defined clinical “triggers” seen earlier in the hospitalization, were more likely to have doc-
that alert the palliative team of a patient meeting criteria umentation of their health care proxy and goals of care, and,
based on certain high-risk populations’ characteristics and out- perhaps most importantly, were less likely to die in the hospital
comes. As an example, one of the first studies using this (Table 4).112 However, the reproducibility or scalability of this
“trigger” approach compared outcomes for patients with kind of approach outside of a single-center with an established
global cerebral ischemia after cardiopulmonary resuscitation palliative medicine program is not known.
(CPR) or multi-organ failure in the ICU. They compared
EOL care in the trigger-based palliative care group versus his- Integrative Model. Despite the proliferation of palliative ser-
torical controls. Length of time between identification of poor vices in hospitals, workforce limitations may preclude consulta-
prognosis and establishment of comfort care goals was tive models for all ICU patients with unmet palliative needs.
decreased in the proactive group as was the amount of time This requires ICUs to also adopt an integrative model, in
dying patients remained in the ICU.108 This approach was which core palliative care principles are incorporated into
also tested among patients with end-stage dementia with routine practice by ICU teams. (Of note, “mixed models” in
similar results –including decreased hospital and MICU which there is a combination of the two approaches, is likely
length of stay, and decreased lag time between identification the most effective and scalable approach at the population
of poor prognosis and establishment of comfort care goals.109 level.113) However, there is no one-size-fits-all approach to
The use of automated or semi-automated clinical “triggers” improve ICU care through palliative interventions.
has been evaluated to identify patients at high risk for unmet Mosenthal and colleagues integrated palliative care into sur-
palliative care needs. Norton et al examined the effect of proac- gical and trauma ICUs through the creation of a six-step
tive palliative care consultation among medical ICU patients pathway for all patients from the time of ICU admission.
deemed high risk for dying (ICU admission after hospital stay Within 24 h, psychosocial services were provided by palliative
>10 days, age >80 with two or more comorbidities, stage IV care counselors or pastoral care with an interdisciplinary palli-
cancer, recent cardiac arrest, or an intracranial hemorrhage ative care needs assessment that evaluated patient symptoms,
requiring mechanical ventilation). Proactive palliative care con- family understanding, and likely prognosis. Within 72 h, a pal-
sultation for these high-risk patients was associated with liative specialist-facilitated family meeting and a comprehen-
decreased medical ICU length of stay.110 In a more recent sive palliative care plan were established. For patients
study, Ma and colleagues examined the effects of early trig- imminently dying, a palliative care standing order set was
gered palliative care consultation on outcomes. Trigger criteria also developed with an accompanying reference sheet.
included at least one of the following criteria: admitted from a Finally, palliative care performance measures were integrated
skilled nursing facility, end stage neurodegenerative condition, into morbidity and mortality conferences. While this pathway
advanced/metastatic cancer, cardiac arrest with neurological created a more standardized approach to ICU palliative care,
compromise, multiple organ system failure, end stage organ individual processes such as symptom assessment or family
disease (decompensated cirrhosis, end stage renal disease, meetings were already being done “as needed”.114 After imple-
class III or IV heart failure, chronic obstructive pulmonary menting this six-step pathway authors found that discussions of
disease on home oxygen), acute shock requiring at least six goals of care on rounds increased, and DNR/comfort measures
hours of vasopressors or inotropic support, acute respiratory were implemented earlier in the hospital course without a
failure requiring invasive or non-invasive ventilation, hospital change in overall mortality rates.115
stay >5 days, or ICU re-admission for the same diagnosis Specialists within the medical ICU have also worked to inte-
within the past 30 days); 199 patients were randomized to grate palliative care into the ICU. Billings and colleagues
receive the palliative trigger intervention or usual care. The embarked on a three-year project where critical care and palli-
patients in the intervention group had a palliative care consulta- ative care clinicians collaborated to improve EOL care in the
tion within 48 h of ICU admission. More patients in the inter- ICU. Following commitment from ICU leaders (“buy-in”) to
vention group changed their code status to DNR (50% vs foster culture change, they identified discrete, concrete areas
23%) and had more transfers to hospice (18.6% vs 4.9%) for collaboration. They identified palliative nurse champions
(Table 4). They also were less likely to have a tracheotomy per- eager to adopt palliative care into their daily practice, provided
formed (1.0% vs 7.8%), had fewer days on the ventilator (4 vs 6 additional palliative care training, and encouraged champions to
Doherty et al. 727

complete their own palliative care-focused project. They also a French trial of sending condolence letters from physicians/
opened the ICU to visitors at all hours. Palliative care specialists nurses in charge at the time of death to SDMs, there was no
were included on rounds and family meetings were encouraged. effect on SDM depression at one month and increased depres-
Finally, a “Get to Know Me” Poster was placed in all patient’s sion and PTSD-related symptoms at 6 months.123
rooms where the family could share information about the A different palliative-related intervention targeting patients
patient’s background, hobbies, and interests.116 While these directly has been the use of ICU diaries. A randomized con-
homegrown process improvement projects are feasible, there trolled trial of ICU patients admitted for > 72 h compared
remained persistent unmet ICU palliative needs (family com- patients given a diary outlining the details of their stay to
munication and support during care transitions). Commitment those who were not given a diary. PTSD was lower in the
from ICU leadership is required to develop nimble, responsive diary compared to the control group.93 However, in a larger
approaches to unit-specific domains most in need to enhance study for ICU patients across 35 French ICUs, a similar inter-
ICU palliative care. vention using diaries failed to improve PTSD-related symp-
Despite encouraging results from individual unit-level toms.94 These findings raise additional questions over which
quality improvement interventions, the ability to develop effec- patients may be most likely to benefit from ICU diaries
tive ICU palliative care delivery models that improve patient during their recovery.
and family-centered outcomes remains elusive (Table 4). In a More recent work has employed other low-cost yet innova-
landmark intervention study (n = 253 pre-intervention vs n = tive approaches for improving EOL care in the ICU. A 3-step
337 post-intervention) including clinician education, local program to support families through the dying process in the
champions empowerment, feedback to physicians, and health- ICU did show decreased prolonged grief (15% vs 21%,
care staff support, family-assessed quality of dying was no dif- p-.035) for families counseled in preparation of and after their
ferent in the post-intervention group. There was, however, a loved one’s death.124 In a program called the “Three
statistical improvement in nurse-assessed quality of dying and Wishes”, investigators identified and sought to fulfill small
a reduction in ICU length of stay.117 In a larger-scale follow-up wishes of dying patients and their families. For example, ICU
randomized study of the same intervention at 12 hospitals, teams identified and honored wishes such as playing the
again the intervention failed to improve the quality of dying, patient’s favorite music, providing a non-hospital blanket, or
ICU length of stay, or time from ICU admission to withdrawal organizing a final celebration with family. Bereaved families
of LSTs. The authors concluded that effective interventions to who participated in Three Wishes reported having received sig-
integrate palliative care principles into critical care and EOL nificantly higher emotional and spiritual support from the
care require more direct contact with patients and families.118 healthcare team compared to those of who did not.125 Given
Intentional communication training efforts for internal medicine the low cost and potential benefit of this sort of intervention,
and advanced practice nurses in the ICU did not improve Three Wishes and similar programs have been more widely
family-assessed general or EOL-specific communication adopted across several ICUs (https://3wishesproject.com/).126
ratings. Of greater concern, depression risk appeared to be
higher in the intervention group, with statistically significantly
higher depression scores in those treated by learners in the com- Persistent Gaps in ICU Palliative Care and Future
munication training sessions.119 Even communication facilita-
tors (rather than training of existing personnel) did not Investigations
improve anxiety or PTSD, although ICU and hospital length To date, efforts to address unmet palliative ICU needs are based
of stay were shorter for the intervention group.120 A similar on results of quality improvement or clinical trial interventions
multidimensional family support intervention from another with modest numbers of patients (n range 52-1866) or partici-
group similarly found no effect on anxiety and depression pants and nearly all from higher income countries with
scores among 1000 SDMs of patients with high likelihood of greater access to healthcare resources (Table 4). Future investi-
death or disability post-ICU.121 gations and recommendations would benefit from identification
Subgroup analyses of decedents from Curtis 2016 and White and routine incorporation of structured data fields to measure
2018 offer some insight into which patients and SDMs might processes of ICU palliative care across large-scale registries
warrant more attention. For both ICU palliative interventions representing multiple countries and cultures.127,128 In more
(communication facilitator and family support), the respective resource-limited settings, including conflict-ridden regions
interventions were associated with decreased ICU and hospital such as civil wars, coups and hostile invasions of sovereign
lengths of stay. Additional evidence from a French study countries, ICU palliative care needs and delivery are likely
offered further hope. Through proactive EOL conferences/ markedly different from the needs and experiences of higher
family meetings and providing a bereavement brochure to fam- income countries not directly in the throes of military con-
ilies, symptoms of PTSD, depression, and anxiety decreased flicts.129,130 In addition to differences in access to resources,
among family members of ICU decedents at 90-day follow-up cultural attitudes, stigma (the “war on drugs” discouraging
compared to individuals receiving usual care.122 On the other use of opioids) and religious beliefs may hinder proactive,
hand, there may be unanticipated hazards related to these inter- effective integration of palliative care into routine practice in
ventions between the treatment teams and bereaved families. In resource-limited settings.129 Future work is needed to expand
728 Journal of Intensive Care Medicine 39(8)

on ICU palliative care needs assessment and optimal delivery 2000-2015. JAMA. 2018;320(3):264. doi:10.1001/JAMA.2018.
models in more resource-limited settings.131 8981
3. Angus DC, Barnato AE, Linde-Zwirble WT, et al. Use of inten-
sive care at the end of life in the United States: An epidemiologic
Conclusions study. Crit Care Med. 2004;32(3):638-43. doi:10.1097/01.CCM.
0000114816.62331.08
Despite progress made in recognizing and addressing palliative 4. Zimmerman JE, Kramer AA, Knaus WA. Changes in hospital
needs of ICU patients, patients still experience high rates of dis- mortality for United States intensive care unit admissions from
tressing and often undertreated symptoms—most commonly 1988 to 2012. Crit Care. 2013;17(2):R81. doi:10.1186/cc12695
pain, dyspnea, and thirst. Communication between physicians, 5. Rai S, Brace C, Ross P, et al. Characteristics and Outcomes of
patients, and families is often of poor quality and can lead to Very Elderly Patients Admitted to Intensive Care: A
significant harm in the high-stakes environment of critical Retrospective Multicenter Cohort Analysis. Crit Care Med.
care. Among patients that survive their ICU stay, many have 51(10):1328-1338. doi:10.1097/CCM.0000000000005943
long-term morbidity related to ongoing physical, psychological, 6. Lone NI, Gillies MA, Haddow C, et al. Five-year mortality
and cognitive complications. Palliative care needs are likely and hospital costs associated with surviving intensive care. Am
J Respir Crit Care Med. 2016;194(2):198-208. doi:10.1164/
unmet in the subset of ICU patients who experience chronic
RCCM.201511-2234OC/SUPPL_FILE/DISCLOSURES.PDF
critical illness with a very high symptom burden and a 50%
7. Chesley CF, Harhay MO, Small DS, Hanish A, Prescott HC,
one-year mortality rate. Mikkelsen ME. Hospital Readmission and Post-Acute Care
The field of palliative care is well poised to address these Use After Intensive Care Unit Admissions: New ICU Quality
challenges in critical care. Many approaches have been taken Metrics? J Intensive Care Med. 2022;37(2):168-176. doi:10.
to integrate palliative care principles into critical care and addi- 1177/0885066620956633
tional work is needed to determine the optimal approach. 8. Hill AD, Fowler RA, Pinto R, Herridge MS, Cuthbertson BH,
Automatic palliative care consults for patients at high risk of Scales DC. Long-term outcomes and healthcare utilization fol-
dying may be a necessary start but are unlikely sufficient. lowing critical illness-a population-based study. Published
Patient- and clinician-facing interventions with the most online 2016. doi:10.1186/s13054-016-1248-y.
promise for improving outcomes for patients and families prior- 9. Detsky ME, Harhay MO, Bayard DF, et al. Six-month morbidity
itize comprehensive symptom assessment, communication and mortality among intensive care unit patients receiving
life-sustaining therapy: A prospective cohort study. Ann Am
early and often and fulfilling feasible patient wishes, including
Thorac Soc. 2017;14(10):1562-1570. doi:10.1513/ANNALSATS.
open exploration of culturally sensitive communication style 201611-875OC/SUPPL_FILE/DISCLOSURES.PDF
preferences of patients/SDMs. Ongoing research with commu- 10. Akgün KM, Krishnan S, Butt AA, et al. HIV-infected medical
nity engagement of patients and SDMs is needed to determine ICU (MICU) survivors without CD4 cell recovery are at
how to optimally integrate palliative and critical care. increased risk for poor outcomes regardless of viral suppression
in a national cohort. AIDS. 2021;35(14):2355. doi:10.1097/
Declaration of Conflicting Interests QAD.0000000000003019
The authors declared no potential conflicts of interest with respect to 11. Herridge MS, Tansey CM, Matté A, et al. Functional disability 5
the research, authorship, and/or publication of this article. years after acute respiratory distress syndrome. N Engl J Med.
2011;364(14):1293-1304. doi:10.1056/NEJMOA1011802/SUPPL_
FILE/NEJMOA1011802_DISCLOSURES.PDF
Funding 12. Dowdy DW, Eid MP, Sedrakyan A, et al. Quality of life in adult
Dr Feder is funded by the Yale Scholars in Implementation Science survivors of critical illness: A systematic review of the literature.
NHLBI K12HL138037 and the National Heart, Lung, and Blood Intensive Care Med. 2005;31(5):611-20. doi:10.1007/s00134-
Institute. National Heart, Lung, and Blood Institute, (grant number 005-2592-6
K12HL138037). The views expressed in this article are the authors’ 13. Aslakson RA, Randall Curtis J, Nelson JE. The changing role of
and do not necessarily reflect the positions or policies of the authors’ palliative care in the ICU. Crit Care Med. 2014;42(11):2418-28.
affiliated Universities and Healthcare systems, nor the Department of doi:10.1097/CCM.0000000000000573
Veterans Affairs or the United States government. 14. Guttormson JL, Khan B, Brodsky MB, et al. Symptom assess-
ment for mechanically ventilated patients: Principles and
ORCID iD priorities: An official American thoracic society workshop
report. Ann Am Thorac Soc. 2023;20(4):491-498. doi:10.1513/
Kathleen M. Akgün MD, MS https://orcid.org/0000-0003-2623-
ANNALSATS.202301-023ST
7463
15. Truog RD, Campbell ML, Curtis JR, et al. Recommendations for
end-of-life care in the intensive care unit: A consensus statement
References by the American college [corrected] of critical care medicine. Crit Care
1. Teno JM, Gozalo PL, Bynum JPW, et al. Change in End-of-life Med. 2008;36(3):953-963. doi:10.1097/CCM.0B013E3181659096
care for medicare beneficiaries: Site of death, place of care, and 16. Lanken PN, Terry PB, DeLisser HM, et al. An official American
health care transitions in 2000, 2005, and 2009. JAMA. thoracic society clinical policy statement: Palliative care for
2013;309(5):470-477. doi:10.1001/JAMA.2012.207624 patients with respiratory diseases and critical illnesses. Am J
2. Teno JM, Gozalo P, Trivedi AN, et al. Site of death, place of care, Respir Crit Care Med. 2008;177(8):912-927. doi:10.1164/
and health care transitions among US medicare beneficiaries, RCCM.200605-587ST
Doherty et al. 729

17. Sullivan DR, Iyer AS, Enguidanos S, et al. Palliative Care Early 32. Schmidt M, Demoule A, Polito A, et al. Dyspnea in mechanically
in the Care Continuum among Patients with Serious Respiratory ventilated critically ill patients. Crit Care Med. 2011;39(9):2059-
Illness An Official ATS/AAHPM/HPNA/SWHPN Policy 2065. doi:10.1097/CCM.0B013E31821E8779
Statement. Am J Respir Crit Care Med. 2022;206(6):e44-e69. 33. Demoule A, Hajage D, Messika J, et al. Prevalence, intensity,
doi:10.1164/rccm.202207-1262ST and clinical impact of dyspnea in critically ill patients receiving
18. Puntillo K, Nelson JE, Weissman D, et al. Palliative care in invasive ventilation. Am J Respir Crit Care Med. 2022;205(8):
the ICU: Relief of pain, dyspnea, and thirst–a report from the 917-926. doi:10.1164/RCCM.202108-1857OC/SUPPL_FILE/
IPAL-ICU advisory board. Intensive Care Med. 2014;40(2): DISCLOSURES.PDF
235-248. doi:10.1007/S00134-013-3153-Z 34. Rotondi AJ, Chelluri L, Sirio C, et al. Patients’ recollections of
19. Nelson JE, Meier DE, Oei EJ, et al. Self-reported symptom expe- stressful experiences while receiving prolonged mechanical ven-
rience of critically ill cancer patients receiving intensive care. tilation in an intensive care unit. Crit Care Med. 2002;30(4):746-
Crit Care Med. 2001;29(2):277-282. doi:10.1097/00003246- 52. doi:10.1097/00003246-200204000-00004
200102000-00010 35. Schmidt M, Banzett RB, Raux M, et al. Unrecognized suffering
20. Devlin JW, Skrobik Y, Gélinas C, et al. Clinical Practice in the ICU: addressing dyspnea in mechanically ventilated
Guidelines for the Prevention and Management of Pain, patients. Intensive Care Med. 2014;40(1):1-10. doi:10.1007/
Agitation/Sedation, Delirium, Immobility, and Sleep Disruption s00134-013-3117-3
in Adult Patients in the ICU. Crit Care Med. 2018;46(9):e825- 36. Gentzler ER, Derry H, Ouyang DJ, et al. Underdetection and
e873. doi:10.1097/CCM.0000000000003299 undertreatment of dyspnea in critically ill patients. Am J Respir
21. Puntillo KA. Pain experiences of intensive care unit patients. Crit Care Med. 2019;199(11):1377-1384. doi:10.1164/RCCM.
Heart and Lung: Journal of Critical Care.1990:19-(5 pt 201805-0996OC/SUPPL_FILE/DISCLOSURES.PDF
1):526-33. 37. Simon ST, Mori M, Ekström M, Pralong A, Yamaguchi T, Hui
22. Gélinas C. Management of pain in cardiac surgery ICU patients: D. Should Benzodiazepines be used for reducing dyspnea in
Have we improved over time? Intensive Crit Care Nurs. patients with advanced illnesses? J Pain Symptom Manage.
2007;23(5):298-303. doi:10.1016/j.iccn.2007.03.002 2023;65(3):e219-e223. doi:10.1016/j.jpainsymman.2022.11.018
23. Puntillo KA, Arai S, Cohen NH, et al. Symptoms experienced 38. Navigante AH, Cerchietti LCA, Castro MA, Lutteral MA,
by intensive care unit patients at high risk of dying. Crit Cabalar ME. Midazolam as adjunct therapy to morphine in the
Care Med. 2010;38(11):2155-2160. doi:10.1097/CCM.0B013E alleviation of severe dyspnea perception in patients with
3181F267EE advanced cancer. J Pain Symptom Manage. 2006;31(1):38-47.
24. Puntillo KA, White C, Morris AB, et al. Patients’ perceptions and doi:10.1016/j.jpainsymman.2005.06.009
responses to procedural pain: Results from thunder project II. Am 39. Stotts NA, Arai SR, Cooper BA, Nelson JE, Puntillo KA.
J Crit Care. 2001;10(4):238-51. doi:10.4037/ajcc2001.10.4.238 Predictors of thirst in intensive care unit patients. J Pain
25. Puntillo KA, Max A, Timsit JF, et al. Determinants of procedural Symptom Manage. 2015;49(3):530. doi:10.1016/J.JPAINSY
pain intensity in the intensive care unit: The Europain® study. MMAN.2014.07.001
Am J Respir Crit Care Med. 2014;189(1):39-47. doi:10.1164/ 40. Landström M, Rehn IM, Frisman GH. Perceptions of registered
rccm.201306-1174OC and enrolled nurses on thirst in mechanically ventilated adult
26. Payen JF, Chanques G, Mantz J, et al. Current practices in seda- patients in intensive care units-a phenomenographic study.
tion and analgesia for mechanically ventilated critically ill patients: Intensive Crit Care Nurs. 2009;25(3):133-139. doi:10.1016/J.
A prospective multicenter patient-based study. Anesthesiology. ICCN.2009.03.001
2007;106(4):687-95. doi:10.1097/01.anes.0000264747.09017.da 41. Nelson JE, Meier DE, Litke A, Natale DA, Siegel RE, Morrison
27. Devlin JW, Roberts RJ. Pharmacology of Commonly Used RS. The symptom burden of chronic critical illness. Crit Care
Analgesics and Sedatives in the ICU: Benzodiazepines, Med. 2004;32(7):1527-1534. doi:10.1097/01.CCM.0000129485.
Propofol, and Opioids. Crit Care Clin. 2009;25(3):431-49. 08835.5A
doi:10.1016/j.ccc.2009.03.003 42. Kjeldsen CL, Hansen MS, Jensen K, Holm A, Haahr A, Dreyer
28. Reintam Blaser A, Preiser JC, Fruhwald S, et al. Gastrointestinal P. Patients’ experience of thirst while being conscious and
dysfunction in the critically ill: a systematic scoping review and mechanically ventilated in the intensive care unit. Nurs Crit
research agenda proposed by the Section of Metabolism, Care. 2018;23(2):75-81. doi:10.1111/NICC.12277
Endocrinology and Nutrition of the European Society of 43. Puntillo K, Arai SR, Cooper BA, Stotts NA, Nelson JE. A ran-
Intensive Care Medicine. Crit Care. 2020;24(1):224. doi:10. domized clinical trial of an intervention to relieve thirst and
1186/s13054-020-02889-4 dry mouth in intensive care unit patients. Intensive Care Med.
29. Oczkowski SJW, Duan EH, Groen A, Warren D, Cook DJ. The 2014;40(9):1295-1302. doi:10.1007/s00134-014-3339-z
use of bowel protocols in critically ill adult patients: A systematic 44. Puntillo KA, Neuhaus J, Arai S, et al. Challenge of assessing
review and meta-analysis. Crit Care Med. 2017;45(7):e718- symptoms in seriously ill intensive care unit patients: Can
e726. doi:10.1097/CCM.0000000000002315 proxy reporters help?*. Crit Care Med. 2012;40(10):2760-
30. Chanques G, Jaber S, Barbotte E, et al. Impact of systematic 2767. doi:10.1097/CCM.0b013e31825b94d8
evaluation of pain and agitation in an intensive care unit. Crit 45. Puntillo KA, Miaskowski C, Kehrle K, Stannard D, Gleeson S,
Care Med. 2006;34(6):1691-1699. doi:10.1097/01.CCM. Nye P. Relationship between behavioral and physiological indi-
0000218416.62457.56 cators of pain, critical care patients’ self-reports of pain, and
31. Kalfon P, Baumstarck K, Estagnasie P, et al. A tailored multi- opioid administration. Crit Care Med. 1997;25(7):1159-1166.
component program to reduce discomfort in critically ill patients: doi:10.1097/00003246-199707000-00017
a cluster-randomized controlled trial. Intensive Care Med. 46. Puntillo KA, Neuhaus J, Arai S, et al. Challenge of assessing
2017;43(12):1829-1840. doi:10.1007/s00134-017-4991-x symptoms in seriously ill intensive care unit patients: Can
730 Journal of Intensive Care Medicine 39(8)

proxy reporters help?*. Crit Care Med. 2012;40(10):2760-2767. 2011;59(11):2091-2099. doi:10.1111/J.1532-5415.2011.

doi:10.1097/CCM.0b013e31825b94d8 03656.X
47. Hanson LC, Danis M, Garrett J. What is wrong with end-of-life 60. Oo NM, Scott IA, Maggacis R, Rajakaruna N. Assessing concor-
care? Opinions of bereaved family members. J Am Geriatr Soc. dance between patient preferences in advance care plans and
1997;45(11):1339-44. doi:10.1111/j.1532-5415.1997.tb02933.x in-hospital care. Aust Health Rev. 2019;43(4):425-431. doi:10.
48. Teno JM, Clarridge BR, Casey V, et al. Family perspectives on 1071/AH18011
end-of-life care at the last place of care. JAMA. 2004;291(1):88- 61. Song MK, Ward SE, Hanson LC, Metzger M, Kim S.
93. doi:10.1001/JAMA.291.1.88 Determining consistency of surrogate decisions and End-of-life
49. Su A, Lief L, Berlin D, et al. Beyond pain: Nurses’ assessment care received with patient goals-of-care preferences. J Palliat
of patient suffering, dignity, and dying in the intensive care unit. Med. 2016;19(6):610. doi:10.1089/JPM.2015.0349
J Pain Symptom Manage. 2018;55(6):1591-1598.e1. doi:10. 62. Chang DW, Neville TH, Parrish J, et al. Evaluation of
1016/J.JPAINSYMMAN.2018.02.005 Time-Limited Trials among Critically Ill Patients with
50. Motamedi M, Brandenburg C, Bakhit M, et al. Concerns and Advanced Medical Illnesses and Reduction of Nonbeneficial
potential improvements in end-of-life care from the perspectives ICU Treatments. JAMA Intern Med. 2021;181(6):786-794.
of older patients and informal caregivers: A scoping review. doi:10.1001/jamainternmed.2021.1000
BMC Geriatr. 2021;21(1):1-12. doi:10.1186/S12877-021- 63. De Miranda S, Pochard F, Chaize M, et al. Postintensive care unit
02680-2/FIGURES/1 psychological burden in patients with chronic obstructive pulmo-
51. Lynn J, Teno JM, Phillips RS, et al. Perceptions by family nary disease and informal caregivers: A multicenter study. Crit
members of the dying experience of older and seriously ill Care Med. 2011;39(1):112-118. doi:10.1097/CCM.0B013E31
patients. SUPPORT investigators. Study to understand progno- 81FEB824
ses and preferences for outcomes and risks of treatments. Ann 64. Pochard F, Darmon M, Fassier T, et al. Symptoms of anxiety and
Intern Med. 1997;126(2):97-106. doi:10.7326/0003-4819- depression in family members of intensive care unit patients
126-2-199701150-00001 before discharge or death. A prospective multicenter study.
52. White DB, Engelberg RA, Wenrich MD, Lo B, Curtis JR. J Crit Care. 2005;20(1):90-96. doi:10.1016/j.jcrc.2004.11.004
Prognostication during physician-family discussions about 65. Siegel MD, Hayes E, Vanderwerker LC, Loseth DB, Prigerson
limiting life support in intensive care units. Crit Care HG. Psychiatric illness in the next of kin of patients who die in
Med. 2007;35(2):442-448. doi:10.1097/01.CCM.0000254723. the intensive care unit. Crit Care Med. 2008;36(6):1722-1728.
28270.14 doi:10.1097/CCM.0B013E318174DA72
53. Hancock K, Clayton JM, Parker SM, et al. Discrepant 66. Anderson WG, Arnold RM, Angus DC, Bryce CL. Posttraumatic
Perceptions About End-of-Life Communication: A Systematic stress and complicated grief in family members of patients in the
Review. J Pain Symptom Manage. 2007;34(2):190-200. doi:10. intensive care unit. J Gen Intern Med. 2008;23(11):1871-1877.
1016/j.jpainsymman.2006.11.009 doi:10.1007/s11606-008-0770-2
54. Witkamp FE, van Zuylen L, Vergouwe Y, van der Rijt CCD, van 67. Johnson CC, Suchyta MR, Darowski ES, et al. Psychological
der Heide A. Concordance between experiences of bereaved rel- sequelae in family caregivers of critically III intensive
atives, physicians, and nurses with hospital End-of-life care: care unit patients. A systematic review. Ann Am Thorac Soc.
Everyone has their “own truth.”. International Journal of 2019;16(7):894-909. doi:10.1513/ANNALSATS.201808-540SR
Palliative Care. 2015;2015 (11):1-10. doi:10.1155/2015/623890 68. Choi JY, Donahoe MP, Zullo TG, Hoffman LA. Caregivers of
55. Unroe KT, Hickman SE, Torke AM. Care consistency with doc- the chronically critically ill after discharge from the intensive
umented care preferences: Methodologic considerations for care unit: Six months’ experience. Am J Crit Care.
implementing the “measuring what matters” quality indicator. 2011;20(1):12-23. doi:10.4037/AJCC2011243
J Pain Symptom Manage. 2016;52(4):453-458. doi:10.1016/J. 69. Cox CE, Docherty SL, Brandon DH, et al. Surviving critical
JPAINSYMMAN.2016.04.015 illness: Acute respiratory distress syndrome as experienced by
56. Andersen SK, Butler RA, Chang CCH, Arnold R, Angus DC, patients and their caregivers. Crit Care Med. 2009;37(10):
White DB. Prevalence of long-term decision regret and associ- 2702-2708. doi:10.1097/CCM.0B013E3181B6F64A
ated risk factors in a large cohort of ICU surrogate decision 70. MacIntyre NR, Epstein SK, Carson S, Scheinhorn D,
makers. Crit Care. 2023;27(1):61. doi:10.1186/S13054-023- Christopher K, Muldoon S. Management of patients requiring
04332-W prolonged mechanical ventilation: Report of a NAMDRC con-
57. Comer AR, Hickman SE, Slaven JE, et al. Assessment of dis- sensus conference. Chest. 2005;128(6):3937-3955. Accessed
cordance between surrogate care goals and medical treatment May 13, 2023. https://go-gale-com.yale.idm.oclc.org/ps/i.do?p=
provided to older adults with serious illness. JAMA Netw Open. HRCA&sw=w&issn=00123692&v=2.1&it=r&id=GALE%
2020;3(5):e205179. doi:10.1001/JAMANETWORKOPEN.2020. 7CA140707342&sid=googleScholar&linkaccess=fulltext.
5179 71. Nelson JE, Cox CE, Hope AA, Carson SS. Chronic critical
58. Hickman SE, Keevern E, Hammes BJ. Use of the physician illness. Am J Respir Crit Care Med. 2010;182(4):446-454.
orders for life-sustaining treatment program in the clinical doi:10.1164/RCCM.201002-0210CI
setting: A systematic review of the literature. J Am Geriatr 72. Kahn JM, Le T, Angus DC, et al. The epidemiology of chronic
Soc. 2015;63(2):341-350. doi:10.1111/JGS.13248 critical illness in the United States. Crit Care Med.
59. Hickman SE, Nelson CA, Moss AH, Tolle SW, Perrin NA, 2015;43(2):282-287. doi:10.1097/CCM.0000000000000710
Hammes BJ. The consistency between treatments provided 73. Gardner AK, Ghita GL, Wang Z, et al. The development of
to nursing facility residents and orders on the physician chronic critical illness determines physical function, quality of
orders for life-sustaining treatment form. J Am Geriatr Soc. life, and long-term survival among early survivors of sepsis in
Doherty et al. 731

surgical ICUs. Crit Care Med. 2019;47(4):566-573. doi:10.1097/ 90. Jones C, Griffiths RD, Humphris G, Skirrow PM. Memory, delu-
CCM.0000000000003655 sions, and the development of acute posttraumatic stress
74. Nelson JE, Hope AA. Integration of palliative care in chronic disorder-related symptoms after intensive care. Crit Care Med.
critical illness management. Respir Care. 2012;57(6):1004. 2001;29(3):573-80. doi:10.1097/00003246-200103000-00019
doi:10.4187/RESPCARE.01624 91. Rabiee A, Nikayin S, Hashem MD, et al. Depressive symptoms
75. Cox CE, Carson SS, Lindquist JH, Olsen MK, Govert JA, after critical illness: A systematic review and meta-analysis. Crit
Chelluri L. Differences in one-year health outcomes and resource Care Med. 2016;44(9):1744-1753. doi:10.1097/CCM.000000
utilization by definition of prolonged mechanical ventilation: A 0000001811
prospective cohort study. Crit Care. 2007;11(1):1-11. doi:10. 92. Hatch R, Young D, Barber V, Griffiths J, Harrison DA, Watkinson
1186/CC5667/FIGURES/5 P. Anxiety, Depression and Post Traumatic Stress Disorder after
76. Carson SS, Garrett J, Hanson LC, et al. A prognostic model for critical illness: a UK-wide prospective cohort study. Crit Care.
one-year mortality in patients requiring prolonged mechanical 2018;22(1):310. doi:10.1186/S13054-018-2223-6
ventilation. Crit Care Med. 2008;36(7):2061. doi:10.1097/ 93. Jones C, Bäckman C, Capuzzo M, et al. Intensive care diaries
CCM.0B013E31817B8925 reduce new onset post traumatic stress disorder following critical
77. Cinotti R, Voicu S, Jaber S, et al. Tracheostomy and long-term illness: A randomised, controlled trial. Crit Care. 2010;14(5):
mortality in ICU patients undergoing prolonged mechanical ven- R168. doi:10.1186/CC9260
tilation. PLoS One. 2019;14(10):e0220399. doi:10.1371/journal. 94. Garrouste-Orgeas M, Flahault C, Vinatier I, et al. Effect of an
pone.0220399 ICU Diary on Posttraumatic Stress Disorder Symptoms Among
78. Nelson JE, Mercado AF, Camhi SL, et al. Communication about Patients Receiving Mechanical Ventilation: A Randomized
chronic critical illness. Arch Intern Med. 2007;167(22):2509. Clinical Trial. JAMA – Journal of the American Medical
doi:10.1001/ARCHINTE.167.22.2509 Association. 2019;322(3):229-239. doi:10.1001/jama.2019.9058
79. Carson SS, Cox CE, Wallenstein S, et al. Effect of palliative 95. Kalfon P, Alessandrini M, Boucekine M, et al. Tailored multi-
care–led meetings for families of patients with chronic critical component program for discomfort reduction in critically ill
illness: A randomized clinical trial. JAMA. 2016;316(1):51-62. patients may decrease post-traumatic stress disorder in general
doi:10.1001/JAMA.2016.8474 ICU survivors at 1 year. Intensive Care Med. 2019;45(2):223-
80. Rawal G, Yadav S, Kumar R. Post-intensive care syndrome: An 235. doi:10.1007/s00134-018-05511-y
overview. J Transl Int Med. 2017;5(2):90-92. doi:10.1515/ 96. Cuthbertson BH, Roughton S, Jenkinson D, Maclennan G, Vale
JTIM-2016-0016 L. Quality of life in the five years after intensive care: a cohort
81. Choi J, Hoffman LA, Schulz R, et al. Self-reported physical study. Crit Care. 2010;14(1):R6. doi:10.1186/cc8848
symptoms in intensive care unit (ICU) survivors: Pilot explora- 97. Jackson JC, Hart RP, Gordon SM, et al. Six-month neuropsycho-
tion over four months post-ICU discharge. J Pain Symptom logical outcome of medical intensive care unit patients. Crit Care
Manage. 2014;47(2):257-270. doi:10.1016/J.JPAINSYMMAN. Med. 2003;31(4):1226-1234. doi:10.1097/01.CCM.0000059996.
2013.03.019 30263.94
82. Boyle M, Murgo M, Adamson H, Gill J, Elliott D, Crawford M. 98. Ehlenbach WJ, Hough CL, Crane PK, et al. Association
The effect of chronic pain on health related quality of life among between acute care and critical illness hospitalization and cog-
intensive care survivors. Aust Crit Care. 2004;17(3):104-113. nitive function in older adults. JAMA : The Journal of the
doi:10.1016/S1036-7314(04)80012-2 American Medical Association. 2010;303(8):763. doi:10.
83. Marra A, Pandharipande PP, Girard TD, et al. Co-occurrence of 1001/JAMA.2010.167
post-intensive care syndrome problems among 406 survivors of 99. Pandharipande PP, Girard TD, Jackson JC, et al. Long-Term
critical illness. Crit Care Med. 2018;46(9):1393. doi:10.1097/ cognitive impairment after critical illness. N Engl J Med.
CCM.0000000000003218 2013;369(14):1306-1316. doi:10.1056/nejmoa1301372
84. Eaton TL, Lewis A, Donovan HS, et al. Examining the needs of 100. Medrinal C, Prieur G, Bonnevie T, et al. Muscle weakness, func-
survivors of critical illness through the lens of palliative care: A tional capacities and recovery for COVID-19 ICU survivors.
qualitative study of survivor experiences. Intensive Crit Care doi:10.1186/s12871-021-01274-0.
Nurs. 2023;75:103362. doi:10.1016/J.ICCN.2022.103362 101. Mongodi S, Salve G, Tavazzi G, Politi P, Mojoli F. High preva-
85. Davydow DS, Gifford JM, Desai S V, Bienvenu OJ, Needham lence of acute stress disorder and persisting symptoms in ICU
DM. Depression in general intensive care unit survivors: A sys- survivors after COVID-19. Intensive Care Med. 2021;47(5):
tematic review. Intensive Care Med. 2009;35(5):796-809. 616-618. doi:10.1007/s00134-021-06349-7
doi:10.1007/S00134-009-1396-5 102. Weidman K, LaFond E, Hoffman KL, et al. Post-Intensive care
86. Weinert C, Meller W. Epidemiology of depression and antide- unit syndrome in a cohort of COVID-19 survivors in
pressant therapy after acute respiratory failure. Psychosomatics. New York city. Ann Am Thorac Soc. 2022;19(7):1158-1168.
2006;47(5):399-407. doi:10.1176/APPI.PSY.47.5.399 doi:10.1513/ANNALSATS.202104-520OC
87. Rattray JE, Johnston M, Wildsmith JAW. Predictors of emo- 103. Parker AJ, Humbir A, Tiwary P, et al. Recovery after critical
tional outcomes of intensive care. Anaesthesia. 2005;60(11): illness in COVID-19 ICU survivors. Br J Anaesth.
1085-1092. doi:10.1111/J.1365-2044.2005.04336.X 2021;126(6):e217-e219. doi:10.1016/J.BJA.2021.03.005
88. Sukantarat K, Greer S, Brett S, Williamson R. Physical and psy- 104. Likhvantsev V, Landoni G, Perekhodov S, et al. Six-Month
chological sequelae of critical illness. Br J Health Psychol. Quality of Life in COVID-19 Intensive Care Unit Survivors. J
2007;12(1):65-74. doi:10.1348/135910706X94096 Cardiothorac Vasc Anesth. 2022;36(7):1949-1955. doi:10.
89. Eddleston JM, White P, Guthrie E. Survival, morbidity, and quality 1053/j.jvca.2021.08.036
of life after discharge from intensive care. Crit Care Med. 105. Dumanovsky T, Augustin R, Rogers M, Lettang K, Meier DE,
2000;28(7):2293-9. doi:10.1097/00003246-200007000-00018 Morrison RS. The growth of palliative care in U.S. Hospitals:
732 Journal of Intensive Care Medicine 39(8)

A status report. J Palliat Med. 2016;19(1):8-15. doi:10.1089/ intensive care unit: A randomized trial. Am J Respir Crit
jpm.2015.0351 Care Med. 2011;183(3):348. doi:10.1164/RCCM.201006-
106. Rogers M, Meier DE, Heitner R, et al. The national palliative 1004OC
care registry: A decade of supporting growth and sustainability 119. Curtis JR, Back AL, Ford DW, et al. Effect of communication
of palliative care programs. J Palliat Med. 2019;22(9):1026- skills training for residents and nurse practitioners on quality
1031. doi:10.1089/JPM.2019.0262 of communication with patients with serious illness: A ran-
107. Feder SL, Jean RA, Bastian L, Akgün KM. National trends in domized trial. JAMA. 2013;310(21):2271-81. doi:10.1001/
palliative care use among older adults with cardiopulmonary jama.2013.282081
and malignant conditions. Heart Lung. 2020;49(4):370. doi:10. 120. Curtis JR, Treece PD, Nielsen EL, et al. Randomized trial of
1016/J.HRTLNG.2020.02.004 communication facilitators to reduce family distress and intensity
108. Campbell ML, Guzman JA. Impact of a proactive approach to of end-of-life care. Am J Respir Crit Care Med. 2016;193(2):
improve end-of-life care in a medical ICU. Chest. 2003;123(1): 154-62. doi:10.1164/rccm.201505-0900OC
266-271. doi:10.1378/CHEST.123.1.266 121. White DB, Angus DC, Shields AM, et al. A Randomized Trial of
109. Campbell ML, Guzman JA. A proactive approach to improve a Family-Support Intervention in Intensive Care Units. N Engl J
end-of-life care in a medical intensive care unit for patients Med. 2018;378(25):2365-2375. doi:10.1056/nejmoa1802637
with terminal dementia. Crit Care Med. 2004;32(9):1839-1843. 122. Lautrette A, Darmon M, Megarbane B, et al. A communication
doi:10.1097/01.CCM.0000138560.56577.88 strategy and brochure for relatives of patients dying in
110. Norton SA, Hogan LA, Holloway RG, Temkin-Greener H, the ICU. N Engl J Med. 2007;356(5):469-478. doi:10.1056/
Buckley MJ, Quill TE. Proactive palliative care in the medical NEJMOA063446/SUPPL_FILE/NEJM_LAUTRETTE_469
intensive care unit: Effects on length of stay for selected high-risk SA1.PDF
patients. Crit Care Med. 2007;35(6):1530-1535. doi:10.1097/01. 123. Kentish-Barnes N, Chevret S, Champigneulle B, et al. Effect of a
CCM.0000266533.06543.0C condolence letter on grief symptoms among relatives of patients
111. Ma J, Chi S, Buettner B, et al. Early palliative care consultation who died in the ICU: a randomized clinical trial. Intensive Care
in the medical ICU: A cluster randomized crossover trial. Crit Med. 2017;43(4):473-484. doi:10.1007/s00134-016-4669-9
Care Med. 2019;47(12):1707-1715. doi:10.1097/CCM.00000 124. Kentish-Barnes N, Chevret S, Valade S, et al. A three-step
00000004016 support strategy for relatives of patients dying in the intensive
112. Mehta A, Krishnasamy P, Chai E, et al. Lessons learned from an care unit: A cluster randomised trial. Lancet. 2022;399(10325):
embedded palliative care model in the medical intensive care 656-664. doi:10.1016/S0140-6736(21)02176-0
unit. J Pain Symptom Manage. 2023;65(4):e321-e327. doi:10. 125. Neville TH, Taich Z, Walling AM, et al. The 3 wishes program
1016/J.JPAINSYMMAN.2022.12.011 improves Families’ experience of emotional and spiritual support
113. Aslakson RA, Curtis JR, Nelson JE. The changing role of palli- at the End of life. J Gen Intern Med. 2023;38(1):115-121. doi:10.
ative care in the ICU HHS public access. Crit Care Med. 1007/S11606-022-07638-7/TABLES/3
2014;42(11):2418-2428. doi:10.1097/CCM.0000000000000573 126. 3 Wishes Project – Helping to make meaningful memories at
114. Mosenthal AC, Murphy PA. Interdisciplinary model for palliative end-of-life. Accessed June 24, 2023. https://3wishesproject.com/
care in the trauma and surgical intensive care unit: Robert Wood 127. Curtis JR, Higginson IJ, White DB. Integrating palliative care
Johnson Foundation Demonstration Project for Improving Palliative into the ICU: a lasting and developing legacy. Intensive Care
Care in the Intensive Care Unit. Crit Care Med. 2006;34(11 Med. 2022;48(7):939-942. doi:10.1007/s00134-022-06729-7
Suppl):S399-403. doi:10.1097/01.CCM.0000237044.79166.E1 128. Ahluwalia SC, Chen C, Raaen L, et al. A Systematic Review in
115. Mosenthal AC, Murphy PA, Barker LK, Lavery R, Retano A, Support of the National Consensus Project Clinical Practice
Livingston DH. Changing the culture around end-of-life care Guidelines for Quality Palliative Care, Fourth Edition. J Pain
in the trauma intensive care unit. Journal of Trauma – Injury, Symptom Manage. 2018;56(6):831-870. doi:10.1016/j.jpainsymman.
Infection and Critical Care. 2008;64(6):1587-1593. doi:10. 2018.09.008
1097/TA.0B013E318174F112 129. Shetty P. The parlous state of palliative care in the developing
116. Billings JA, Keeley A, Bauman J, et al. Merging cultures: world. Lancet. 2010;376(9751):1453-4. doi:10.1016/S0140-
Palliative care specialists in the medical intensive care unit. 6736(10)61978-2
Crit Care Med. 2006;34(11 SUPPL.):S388-93. doi:10.1097/01. 130. Efstathiou N, Vanderspank-Wright B, Vandyk A, et al. Terminal
CCM.0000237346.11218.42 withdrawal of mechanical ventilation in adult intensive care
117. Curtis JR, Treece PD, Nielsen EL, et al. Integrating palliative and units: A systematic review and narrative synthesis of perceptions,
critical care: Evaluation of a quality-improvement intervention. experiences and practices. Palliat Med. 2020;34(9):1140-1164.
Am J Respir Crit Care Med. 2008;178(3):269-275. doi:10. doi:10.1177/0269216320935002
1164/RCCM.200802-272OC/SUPPL_FILE/ 131. Abu-Odah H, Molassiotis A, Liu J. Lessons Learned from
BLUE-200802-272OC_ONLINE_DATA_SUPPLEMENT.PDF Countries That Have Introduced Palliative Care Services into
118. Curtis JR, Nielsen EL, Treece PD, et al. Effect of a Their National Health System: A Narrative Review. J Palliat
quality-improvement intervention on End-of-life care in the Care. 2022;37(1):55-66. doi:10.1177/0825859721989557