Guidelines for the Management of Adult Acute

and Acute-on-Chronic Liver Failure in the ICU:

Neurology, Peri-Transplant Medicine, Infectious
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

Disease, and Gastroenterology Considerations

wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

Rahul Nanchal, MD, MS, FCCM

OBJECTIVES: To develop evidence-based recommendations for clinicians car- (Co-Chair)1
ing for adults with acute liver failure (ALF) or acute on chronic liver failure (ACLF) Ram Subramanian, MD, FCCM
in the ICU. (Co-Chair)2
Waleed Alhazzani, MBBS, MSc,
DESIGN: The guideline panel comprised 27 members with expertise in aspects of FRCPC (Methods Chair, Vice
care of the critically ill patient with liver failure or methodology. We adhered to the Co-Chair)3
Society of Critical Care Medicine standard operating procedures manual and conflict- Joanna C. Dionne, MD, MS, BN3
of-interest policy. Teleconferences and electronic-based discussion among the panel, William J. Peppard, PharmD,
as well as within subgroups, served as an integral part of the guideline development. BCPS, FCCM4
Kai Singbartl, MD, MPH, EDIC,
INTERVENTIONS: In part 2 of this guideline, the panel was divided into four FCCM5
subgroups: neurology, peri-transplant, infectious diseases, and gastrointestinal Jonathon Truwit, MD, MBA6
groups. We developed and selected Population, Intervention, Comparison, and Ali H. Al-Khafaji, MD, MPH, FCCM7
Outcomes (PICO) questions according to importance to patients and practic- Alley J. Killian, PharmD, BCPS2
ing clinicians. For each PICO question, we conducted a systematic review and Mustafa Alquraini, MBBS, SBEM,
meta-analysis where applicable. The quality of evidence was assessed using ABEM, MMed, CCM8
the Grading of Recommendations Assessment, Development, and Evaluation Khalil Alshammari, MBBS8
approach. We used the evidence to decision framework to facilitate recommenda- Fayez Alshamsi, MBBS8
tions formulation as strong or conditional. We followed strict criteria to formulate Emilie Belley-Cote, MD3
best practice statements. Rodrigo Cartin-Ceba, MD, MS9
MEASUREMENTS AND MAIN RESULTS: We report 28 recommendations Steven M. Hollenberg, MD, FACC,
FCCM, FAHA, FCCP10
(from 31 PICO questions) on the management ALF and ACLF in the ICU. Overall,
five were strong recommendations, 21 were conditional recommendations, two Dragos M. Galusca, MD, FASA,
FCCP11
were best-practice statements, and we were unable to issue a recommendation David T. Huang, MD, MPH, FCCM7
for five questions due to insufficient evidence.
Robert C. Hyzy, MD, MCCM12
CONCLUSIONS: Multidisciplinary, international experts formulated evi- Mats Junek, BSc(H), MD3
dence-based recommendations for the management ALF and ACLF patients in Prem Kandiah, MD2
the ICU, acknowledging that most recommendations were based on low quality Gagan Kumar, MD, MA, MS13
and indirect evidence. Rebecca L. Morgan, PhD, MPH14
KEY WORDS: acute liver failure; acute on chronic liver failure; clinical practice Peter E. Morris, MD15
guidelines; Grading of Recommendations Assessment, Development, and Jody C. Olson, MD16
Evaluation Rita Sieracki, MLS6
Randolph Steadman, MD17

I
Beth Taylor, DCN, RDN-AP, CNSC,
n a previous document, we published recommendations for the manage- FCCM18
ment of the critically ill patient with liver disease focused on cardiovas- Constantine J. Karvellas, MD, MS,
FRCPC, FCCM (Vice Co-Chair)19
cular, hematological, pulmonary, renal, and endocrine/nutrition issues (1).
In continuation of the previous document, the current article addresses infec-
tious disease, peri-transplant, gastrointestinal, and neurologic issues that pre-
sent unique challenges in this population of patients. Copyright © 2023 by the Society of
Patients with acute liver failure (ALF) or acute on chronic liver failure Critical Care Medicine.
(ACLF) are at high risk of developing critical illness. Once critical illness DOI: 10.1097/CCM.0000000000005824

Critical Care Medicine www.ccmjournal.org     657

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

occurs, mortality is exceedingly high and often the The panel had a total of 27 members and was then
definitive treatment is liver transplantation (LT). The divided into the following groups: neurology, peri-
unique pathophysiology of liver disease leading to transplant, infectious diseases, and gastrointestinal
critical illness portends unique manifestations in var- groups. Each group was assigned a group leader, a
ious organ systems. Strategies used to manage organ methodologist, and expert panel members. The group
complications in general critical illness are not always leader was responsible for development of Population,
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

applicable to the care of the patient with liver failure. Intervention, Comparison, and Outcomes (PICO)
As with many other illnesses, early recognition and questions for their respective group (with input from
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

prompt management of liver failure and its complica- the chairs and entire guideline committee), leading
tions may improve outcomes. group meetings, assignment of tasks to group mem-
In this document, we provide evidence-based rec- bers, managing activities culminating in recommenda-
ommendations intended to guide the practicing tions, and finalizing drafts of recommendations prior
clinicians (critical care and emergency physicians, to guideline committee voting.
pharmacists, nurses, advanced practice providers, and
dietitians) caring for the critically ill patient with ALF Management of Conflict of Interest
or ACLF. These guidelines are meant to supplement
and not replace an individual clinician’s cognitive de- The guideline panel completed a standardized SCCM
cision-making. The primary goal of these guidelines is conflicts of interest (COI) declaration form. The chairs
to aid best practice and not represent standard of care. of the guideline reviewed and adjudicated all reported
For the purposes of this guideline, we defined ACLF COI by panel members. Individuals who disclosed a
as a syndrome characterized by acute decompensation COI or potential COI (electronically or verbally) dur-
of liver cirrhosis, organ dysfunction, and high short- ing the process of guideline development were asked to
term mortality (2). Presence of organ failure distin- abstain from voting on recommendations where con-
guishes ACLF from acute decompensation of cirrhosis flict existed. The committee followed all procedures
(acute development of ascites, variceal bleeding, and as documented in the American College of Critical
hepatic encephalopathy). In contrast, we defined ALF Care Medicine/SCCM Standard Operating Procedures
by the occurrence of encephalopathy and hepatic syn- Manual. Overall, 11 panel members disclosed poten-
thetic dysfunction within 26 weeks of the first symp- tial secondary COI (intellectual COI). All panel mem-
toms of liver disease in a patient without evidence of bers were asked to disclose any financial COI; none
chronic liver disease (3). disclosed any financial COI. We assigned panel mem-
bers with potential intellectual COI to groups where
COI did not exist.
METHODOLOGY
Selection and Organization of Committee Question Development and Outcome
Members Prioritization
Co-chairs and co-vice chairs were appointed by the In this document, we only included questions from
guidelines committee of the Society of Critical Care four groups (neurology, peri-transplant medicine,
Medicine (SCCM). Chairs and vice chairs in collabo- infectious diseases, and gastrointestinal groups). All
ration with SCCM chose committee members from questions were developed in the PICO format when
two groups of individuals: 1) practicing clinicians applicable. Questions were developed via in-person
with expertise in aspects of care of the critically ill pa- meetings, emails, and teleconferences with input from
tient with liver failure and 2) experts in methodology. the guideline committee. Final decisions regarding
Methodologists were provided by the Guidelines in question inclusion were determined by arriving at con-
Intensive Care, Development, and Evaluation group. sensus through discussion between the co-chairs, vice
Members of the guideline committee were intensivists, chairs, group heads, and methodologists; prioritiza-
gastroenterologists, hepatologists, anesthesiologists, in- tion was based on potential importance to patients and
fectious disease specialists, transplant physicians, phar- end users of the guidelines rather than experts’ per-
macists, dieticians, and advanced practice providers. spectives or interests. While additional questions were

658     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

considered, 32 questions are included in these guide- more studies. For meta-analysis of RCT data, we used
lines. We provide the complete list of PICO questions random-effects model and inverse variance method to
for this document in Supplementary Table 1 (http:// pool estimates across relevant studies. We reported rel-
links.lww.com/CCM/H302). ative risks (RRs) and 95% CI for binary outcomes, and
We used the Grading of Recommendations mean difference and 95% CI for continuous outcomes.
Assessment, Development, and Evaluation (GRADE) For observational (nonrandomized) data, we conducted
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

approach to prioritize outcomes and took the patient meta-analysis if all individual studies provided adjusted
perspective during the prioritization process. First, we estimates and not just crude values and included both an
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

asked panel members in each group to list potentially intervention and a control arm, we used random-effects
relevant outcomes for each PICO questions. Then, we model and inverse variance method to pool adjusted
sent an electronic survey asking each panelist to rate odds ratio (OR) across relevant studies, presenting OR
each of the listed outcomes on a scale from 1 (not im- and 95% CI for binary outcomes. All analyses were
portant) to 9 (critical). Outcomes with a mean rating of conducted using RevMan software (Review Manager,
7 or more were considered critical and were included Version 5.3. Copenhagen, Denmark: The Nordic
under each question. Cochrane Center, The Cochrane Collaboration, 2014).

Systematic Review Grading of Recommendations

For each of the questions, the medical librarian, with The GRADE approach principles guided the assess-
input from panelist and methodologist, performed in- ment of quality of evidence from high to very low and
dependent literature searches. Group members in con- were used to determine the strength of recommenda-
cert with group heads and methodology leads provided tions. The GRADE approach to assess the quality of
pertinent search terms and appropriate key words evidence is based on the evaluation of six domains:
for each question. A minimum of two major data- 1) risk of bias, 2) inconsistency, 3) indirectness, 4)
bases (Medline, Cochrane Registry, or EMBASE) were imprecision, 5) publication bias, and 6) other crite-
searched for relevant studies from inception to 2018. ria (6). The methodologist in each group performed
the initial assessment of quality of evidence (as high,
moderate, low, or very low), incorporated feedback
Screening and Data Abstraction
from panel members, and generated evidence profiles
After finalizing the searches for each PICO question, a using GRADE pro GDT software (Evidence Prime,
panel member screened the titles and abstracts, reviewed Hamilton, ON, Canada) (7).
full text of potentially relevant articles. The aim was to
identify recently published systematic reviews, relevant Formulation of Recommendations
randomized controlled trials (RCTs), and lastly, rele-
vant observational studies. Panel members then used a In a series of webinars, methodologists reviewed the
standardized data abstraction sheet to abstract data on relevant data for each PICO question with subgroup
population, interventions, and outcomes. members to formulate initial recommendations. Each
of the groups used the evidence-to-decision (EtD)
framework to facilitate transition from evidence to
Risk of Bias Assessment
the final recommendation. The EtD framework ensure
Panel members, with input from methodologists, used that panel members take into consideration the quality
the Cochrane risk of bias tool to assess the risk of bias of evidence, magnitude of effect, patients’ values and
of RCTs (4), and Newcastle Ottawa Scale to assess risk preferences, resources, cost, acceptability, and feasi-
of bias of nonrandomized studies (5). bility (8).
Applying the GRADE approach, we classified recom-
Summarizing the Evidence mendations as strong or conditional using the language
“We recommend…” or “We suggest…,” respectively.
When applicable, the methodologists used meta-ana- The strength of a recommendation reflects the confi-
lytic techniques to generate pooled estimates for two or dence regarding whether the desirable consequences

Critical Care Medicine www.ccmjournal.org     659

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

of the recommended intervention would outweigh the Rationale. Cerebral edema is common in ALF, es-
undesirable consequences. Thus, a strong recommen- pecially in patients with grade III and IV hepatic en-
dation in favor of an intervention reflects that the de- cephalopathy. We did not identify any RCTs evaluating
sirable effects of adherence will clearly outweigh the invasive ICP monitoring in patients with ALF. Three
undesirable effects. The implications of calling a recom- observational studies evaluated epidural, subdural and
mendation strong are that most patients would accept intra parenchymal ICP monitors and compared them
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

that intervention and that most clinicians should use it to a control group (frequent neurologic examinations
in most situations. However, a strong recommendation and imaging as needed). Two studies compared mor-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

does not imply a standard of care, and circumstances tality in those who received ICP monitoring versus
may exist in which a strong recommendation cannot or those who did not (11, 12). Bleeding rates were higher
should not be followed for an individual patient. A con- with subdural and intra-parenchymal devices in com-
ditional recommendation indicates that the desirable parison to extradural devices (11–13). The rates of
effects of adherence will probably outweigh the undesir- infection were lowest in extradural devices when com-
able effects, but confidence is diminished either because pared with subdural and intra-parenchymal devices
the quality of evidence or the benefits and risks were (12). However, ICP monitoring was not associated
closely balanced. We anticipate that a conditional rec- with any tangible benefits in outcomes (OR for mor-
ommendation, while still relevant for most patients in tality, 1.21; 95% CI, 0.84–1.75; Supplementary Table
most settings, will be more heavily influenced by clinical 2, http://links.lww.com/CCM/H302). Risk of bias was
circumstances and patients’ values (Table 1). Strong rec- high secondary to the observational nature of the stud-
ommendations based on low quality of evidence can be ies; thus, a conditional recommendation was issued.
justified rarely, such as in life- threatening scenarios or
when there is a critical imbalance in benefit and risk (9). Plasma Exchange for Treatment of Hyperammonemia
Best practice statements (BPSs) were developed as in ALF.
ungraded strong recommendations in adherence with Recommendation. We suggest, when available, using
strict conditions (10). plasma exchange in critically ill ALF patients who de-
velop hyperammonemia (Conditional recommenda-
Voting Process tion, low quality of evidence).
Remarks. Hyperammonemia is defined as ammonia
After each group formulated draft recommendations, level greater than 150 umol/L.
all committee members received links to an electronic Rationale. Hyperammonemia is associated with ce-
survey, each nonconflicted member had to indicate
rebral edema and intracranial hypertension in ALF
agreement or disagreement, while conflicted members
patients. Various modalities have been studied in liter-
abstained from voting on recommendations in which
ature for chronic liver failure; however, there are very
COI exists. We defined consensus and accepted the
limited studies in ALF population. Unlike ACLF, ALF
recommendation if there was 80% consensus agree-
patients are not preconditioned to cope with hyper-
ment among at least 75% of the committee members.
ammonemia and are more susceptible to intracranial
Disagreements were resolved through teleconference
hypertension. Treatments such as lactulose and rifaxi-
calls, emails and revoting with modifications to state-
min used in ACLF, have not demonstrated benefit in
ments to reach consensus. We used up to three rounds
ALF (14–20). Bernal et al (21) evaluated the relation
of voting to resolve disagreements.
of the admission arterial ammonia concentration and
other clinical variables with the development of HE
Neurology Section
and ICH. Variables associated with intracranial hy-
Intracranial Pressure Monitoring. pertension and hepatic encephalopathy were investi-
Recommendation. We suggest not using invasive in- gated; ammonia was an independent risk factor for the
tracranial pressure (ICP) monitoring for critically ill development of both intracranial hypertension and
ALF patients with advanced-grade encephalopathy hepatic encephalopathy. Intracranial hypertension
(Conditional recommendation, very low quality of developed in 55% of ALF patients with an ammonia
evidence). level greater than 200 umol/L (21). Continuous renal

660     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

TABLE 1.
Implications of the Strength of Recommendation
Stakeholder Strong Recommendation Conditional Recommendation

Patients Most individuals in this situation would want the The majority of individuals in this situation
recommended course of action and only a small would want the suggested course of action
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

proportion would not. but many would not.

Clinicians Most individuals should receive the recommended course Different choices are likely to be appropriate
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

of action. Adherence to this recommendation accord- for different patients, and therapy should
ing to the guideline could be used as a quality criterion be tailored to the individual patient’s
or performance indicator. Formal decision aids are not circumstances. Those circumstances may
likely to be needed to help individuals make decisions include the patient or family’s values and
consistent with their values and preferences. preferences.
Policy The recommendation can be adapted as policy in most Policy making will require substantial debates
makers situations including for the use as performance and involvement of many stakeholders.
indicators. Policies are also more likely to vary between
regions. Performance indicators would
have to focus on the fact that adequate
deliberation about the management options
has taken place.

replacement therapy remains the first-line treatment Rationale. In a single-center RCT, Murphy et al
for hyperammonemia and is often used in the absence (23) examined the effect of induced hypernatremia
of acute kidney injury (AKI). Further, ICH in ALF is on the occurrence rate of intracranial hypertension in
driven by both hyperammonemia and systemic in- patients with ALF. Thirty patients with ALF and grade
flammatory response syndrome. High-volume plasma III or IV encephalopathy were randomized. Patients
exchange (HVPE) was found to have a beneficial effect in group 1 (n = 15) received the normal standard of
in one RCT (22). In 92 patients receiving HVPE, care, patients in group 2 (n = 15) received standard
compared with standard medical therapy alone (90 care and hypertonic saline (30%) via infusion to
patients), HVPE improved the LT-free survival rate of maintain serum sodium levels of 145–155 mmol/L.
patients with ALF and grade II hepatic encephalop- ICP was monitored in all patients with a subdural
athy. This amelioration appears to be mainly related to catheter for up to 72 hours after inclusion. Serum so-
the improvement of arterial pressure, with decreased dium levels became significantly different from the
vasopressor requirement. The improvement of the levels observed in the control group at 6 hours. ICP
hospital survival seemed to be limited to the improved decreased significantly relative to baseline over the
outcome of the 68 nontransplanted patients managed first 24 hours in the treatment group but not in the
with HVPE; on meta-analysis, overall risk of mortality control group. The occurrence rate of intracranial
was no different between groups (RR, 0.79; 95% CI, hypertension (ICP > 25 mm Hg or greater) was sig-
0.58–1.08; Supplementary Table 3a, http://links.lww. nificantly higher in the control group. Mortality from
com/CCM/H302). intracranial hypertension was no different between
Therapies to Decrease ICP in Patients With ALF. group (RR, 0.67; 95% CI, 0.13–3.44; Supplementary
Recommendation. We suggest using hypertonic Table 3b, http://links.lww.com/CCM/H302). Rise in
saline in critically ill ALF patients who are at risk of serum sodium levels should be gradual to provide a
developing intracranial hypertension (Conditional constant gradient between brain and plasma. Thirty
recommendation, low quality of evidence). percent saline is not routinely available; thus, in clin-
Remarks. Risk factors for intracranial hypertension ical practice infusions of 3% saline can be used to
include hyperammonemia (> 150 umol/L), high-grade raise sodium levels. Serum sodium levels should be
hepatic encephalopathy or evidence of multiple organ maintained between 145 and 155 mmol/L as dictated
failure (21). by the clinical situation.

Critical Care Medicine www.ccmjournal.org     661

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

Targeted Temperature Management in ALF. reduce hepatic encephalopathy (24 RCTs [n = 1,487];
Recommendation. We suggest not routinely using RR, 0.58; 95% CI, 0.5–0.69) and serious liver-related
induced moderate hypothermia (< 34°C) for critically adverse events such as liver failure, variceal bleed-
ill ALF patients who are at risk of developing intracra- ing, serious infections, spontaneous bacterial peri-
nial hypertension (Conditional recommendation, very tonitis (SBP), and hepatorenal syndrome (24 RCTs
low quality of evidence). [n = 1,487]; RR, 0.47; 95% CI, 0.36–0.6). Treatment
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

Rationale. Moderate hypothermia has been success- was also associated with a reduction in mortality in
ful in decreasing ICP and has been reported to help to patients with overt encephalopathy (RR, 0.36; 95% CI,
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

bridge to liver transplant in some uncontrolled stud- 0.14–0.94; Supplementary Table 4, http://links.lww.
ies (24–26). Its use in ALF remains controversial, as com/CCM/H302), although not in patients with min-
two studies have demonstrated both absence of benefit imal hepatic encephalopathy. The quality of evidence
and harm (27, 28). A retrospective cohort study of ALF was downgraded because the population studied was
patients in the U.S. Acute Liver Failure Study Group cirrhotics with hepatic encephalopathy, and most tri-
with grade III or IV hepatic encephalopathy found that als were at high risk of bias for lack of blinding. Thus, a
therapeutic hypothermia in ALF was not associated conditional recommendation was issued.
with increased bleeding or infections. Although young Recommendation. We suggest using enteral poly-
acetaminophen ALF patients may benefit, therapeutic ethylene glycol (PEG) as an alternative to lactulose in
hypothermia did not consistently affect 21-day sur- critically ill ACLF patients with overt hepatic encepha-
vival (28). In a multicenter RCT (n = 46), patients with lopathy (Conditional recommendation, low quality of
ALF, high-grade encephalopathy, and ICP monitoring evidence).
were randomized to targeted temperature manage- Rationale. A single center RCT (n = 50) demon-
ment groups of 34°C or 36°C (control) for a period of strated that using 4 L of PEG enterally over 4 hours led
72 hours. The primary outcome was a sustained eleva- to faster hepatic encephalopathy resolution compared
tion in ICP greater than 25 mm Hg. There were no sig- with standard therapy with lactulose (30). Thirteen of
nificant differences between the groups in the primary 25 patients in the standard therapy arm (52%) had an
outcome during the study period (35% vs 27%; p = improvement of one or more in hepatic encephalop-
0.56) (RR, 1.31; 95% CI, 0.53–3.2). Furthermore, both athy score, compared with 21 of 23 evaluated patients
groups had similar occurrence rate of adverse events receiving PEG (91%) (RR, 0.18; 95% CI, 0.04–0.72;
and overall mortality (41% vs 46%; p = 0.75; RR, 0.89; Supplementary Table 5, http://links.lww.com/CCM/
95% CI, 0.44–1.80; Supplementary Table 3c, http:// H302). The median time for hepatic encephalopathy
links.lww.com/CCM/H302). This study did not con- resolution was 2 days for standard therapy and 1 day
firm an advantage of induced moderate hypothermia for PEG group. PEG safety profile and balanced elec-
in patients with ALF (29). trolytes make it an attractive alternative to lactulose in
Treatment of Hepatic Encephalopathy. the ICU setting. However, volume of 4 L may be a con-
Recommendation. There was insufficient evidence cern for aspiration, especially in advanced grades of
to issue a recommendation on using lactulose, rifaxi- encephalopathy and should be used cautiously.
min, flumazenil, branch-chain amino acids, carnitine, Recommendation. We suggest using oral rifaxi-
zinc, probiotics, and L-ornithine L-aspartate (LOLA) min as adjunctive therapy in critically ill patients
in critically ill ALF patients with hyperammonemia. ACLF patients with overt hepatic encephalop-
Recommendation. We suggest using nonabsorbable athy (Conditional recommendation, low quality of
disaccharides in critically ill ACLF patients with overt evidence).
hepatic encephalopathy (Conditional recommenda- Rationale. Rifaximin is an oral nonsystemic antibi-
tion, low quality of evidence). otic with less than 0.4% absorption. In a RCT (n = 120)
Rationale. Nonabsorbable disaccharides (NADs) comparing rifaximin (550 mg bid) and lactulose with
(i.e., lactulose, lactitol) are used as first-line agents for lactulose and placebo in which 80% of patients had
the treatment of hepatic encephalopathy. In a meta- severe hepatic encephalopathy, patients who received
analysis of 38 RCTs (n = 1,828), Gluud et al (14) found rifaximin demonstrated an increased proportion of
that NADs, compared with placebo/no intervention, complete encephalopathy reversal and improvement

662     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

in 10-day mortality. In a recent meta-analysis evalu- performance on the number connection test (standard-
ating the role of rifaximin in hepatic encephalopathy ized mean difference, –0.62; 95% CI, –1.12 to –0.11)
(19 RCTs, n = 1,370), rifaximin was associated with a but not in a reduction in encephalopathy recurrence
beneficial effect in secondary prevention of enceph- (RR, 0.64; 95% CI, 0.26–1.59). However, mortality,
alopathy (RR, 1.32; 95% CI, 1.06–1.65) (15). Patients liver-related morbidity, and quality of life were not re-
receiving rifaximin were more likely to recover from ported (31). GPB lowers ammonia by providing an al-
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

hepatic encephalopathy (RR, 0.59; 95% CI, 0.46–0.76) ternate pathway to urea for waste nitrogen excretion in
and had reduced mortality (RR, 0.50; 95% CI, 0.31– the form of phenylacetylglutamine (PAGN), which is
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

0.82; Supplementary Table 6, http://links.lww.com/ excreted in urine. In a randomized phase II trial of 178
CCM/H302). The high cost of rifaximin may be a sig- cirrhotic patients (n = 59 receiving rifaximin) who had
nificant barrier to its routine use. experienced greater than or equal to 2 hepatic enceph-
Recommendation. We suggest using LOLA in criti- alopathy events in the previous 6 months, GPB was as-
cally ill ACLF patients with overt hepatic encephalop- sociated with decreased encephalopathy events, serum
athy (Conditional recommendation, very low quality ammonia levels, and no difference in adverse events
of evidence). (32). GPB use may be cost prohibitive and is limited by
Rationale. LOLA is substrate for urea cycle and its dependence on renal clearance to eliminate PAGN
stimulates enzymatic activity in residual hepatocytes and must be used with caution in the setting of AKI.
leading to increased urea excretion. LOLA is more Acarbose, an alpha-glycosidase inhibitor and hypo-
frequently used for treatment of hepatic encephalop- glycemic agent was tested in one RCT in patients with
athy outside the United States. A recent systematic grade I or II hepatic encephalopathy and type II dia-
review (six RCTs, n = 597) suggested a possible bene- betes. Although there was a salutary effect on serum
ficial effect of LOLA on mortality, hepatic encephalop- ammonia levels, sample size was small, an indirect and
athy, and serious adverse events in comparisons with inaccurate marker of hepatic encephalopathy was used
placebo or no intervention (Supplementary Table and other outcomes such as mortality were not re-
7, http://links.lww.com/CCM/H302) (18). However, ported (33). Please see Supplementary Table 8 (http://
because the quality of the evidence was very low, the links.lww.com/CCM/H302) for complete evidence
panel was very uncertain about these findings. profiles and summary of judgments.
Recommendation. We suggest not routinely using IV
flumazenil, probiotics, zinc supplementation, glycerol
Infectious Diseases
phenylbutyrate (GPB), or acarbose as adjunctive ther-
apies in critically ill ACLF patients with overt hepatic Antibiotic Prophylaxis With Upper Gastrointestinal
encephalopathy (Conditional recommendation, very Bleeding.
low quality of evidence). Recommendation. We recommend using antibiotic
Rationale. A recent systematic review (12 controlled prophylaxis in critically ill ACLF patients with any
trials, n = 842) found low-quality evidence suggesting type of upper gastrointestinal bleeding (UGIB) (Strong
a short-term beneficial effect of IV flumazenil in he- recommendation, moderate quality of evidence).
patic encephalopathy in cirrhosis with no difference Rationale. UGIB is a major risk factor for the subse-
in all-cause mortality (18). If used, flumazenil should quent development of bacterial infections with 45% to
be used in a closely monitored environment as it has a 66% of patients developing infections within the first
potential of provoking seizures. A meta-analysis that 7 days of the bleeding episode. Administration of pro-
included 21 RCTs (n = 1,420) suggested that probiotics phylactic antibiotics (typically third-generation cepha-
may lead to improvements in the development of overt losporins) in ACLF patients with UGIB may attenuate
hepatic encephalopathy (10 RCTs [n = 585; RR, 0.29; the occurrence rate of infections and rebleeding as well
95% CI, 0.16–0.51]). Conversely, probiotics were not as improve survival.
associated with differences in mortality (seven RCTs A meta-analysis of 12 RCTs (n = 1,241) found that an-
[n = 404; RR, 0.58; 95% CI, 0.23–1.44]) (16). Oral tibiotic prophylaxis of bacterial infections in cirrhotic
zinc supplementation from a meta-analysis of four patients with UGIB in comparison to no antibiotic
RCTs (n = 233) showed significant improvement in prophylaxis/placebo was associated with a reduction

Critical Care Medicine www.ccmjournal.org     663

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

in all-cause mortality (RR, 0.79; 95% CI, 0.63–0.98), Remarks. Risk factors for invasive fungal infec-
bacterial infections (RR, 0.35; 95% CI, 0.26–0.47), tions include renal failure requiring dialysis, rejection
bacteremia (RR, 0.25; 95% CI, 0.15–0.40), overall treatment, cytomegalovirus viremia or disease, acute
rebleeding episodes (RR, 0.53; 95% CI, 0.38–0.74), hepatic insufficiency, early graft failure, retransplanta-
and SBP (RR, 0.45; 95% CI, 0.27–0.75) (34). Further tion, preoperative use of broad-spectrum antibiotics,
rebleeding at 7 days was also significantly reduced fungal colonization, and re-exploration after trans-
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

(RR, 0.24; 95% CI, 0.12–0.50). We downgraded the plantation (36).

evidence as included studies were at high risk of bias Rationale. Invasive fungal infections are an impor-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

from lack of blinding and proper sample size calcula- tant cause of mortality and morbidity in liver trans-
tions (Supplementary Table 9, http://links.lww.com/ plant recipients. The most common infections are with
CCM/H302). Candida, followed by Aspergillus. Systemic antifungal
prophylaxis may reduce the occurrence rate of invasive
Albumin Infusion in SBP.
fungal infections and improve outcomes. Conversely,
Recommendation. We recommend using albumin
prophylaxis may also be associated with unnecessary
in critically ill ACLF patients with SBP (Strong recom-
drug toxicity, development of resistance and increased
mendation, moderate quality of evidence).
costs. In a meta-analysis, Evans et al (37) found that sys-
Rationale. SBP is the most common infection-related
temic antifungal prophylaxis compared with placebo
complication in cirrhotic patients with ascites. Once SBP
was associated with a significantly reduced risk of in-
develops, the inherent vasodilated and immune-dys- vasive fungal infections (OR, 0.37; 95% CI, 0.19–0.72)
functional state of cirrhotic patients places them at high and mortality attributable to invasive fungal infections
risk of developing shock, AKI and other organ failures (OR, 0.32; 95% CI, 0.10–0.83). However, overall mor-
(ACLF). In a meta-analysis of four RCTs (288 patients tality was not impacted by the use of prophylaxis (OR,
with SBP), albumin reduced the odds of mortality (OR, 0.87; 95% CI, 0.54–1.39). We downgraded the strength
0.34; 95% CI, 0.19–0.60) and renal impairment (OR, of evidence and issued a conditional recommendation
0.21; 95% CI, 0.11–0.42) (35). Only three trials used no because most included studies were at high risk of bias
albumin as the comparator, while one used an artificial due to small sample sizes, unclear allocation conceal-
colloid. Patients in all four trials received antibiotics. We ment, and inadequate blinding (Supplementary Table
downgraded the evidence based on the lack of blind- 11, http://links.lww.com/CCM/H302).
ing in trials (Supplementary Table 10, http://links.lww. Although risk of acquiring invasive fungal infec-
com/CCM/H302). We issued a strong recommendation tions was attenuated, overall mortality was unchanged.
based on direct evidence of the application of albumin in Weighing the risks versus benefits, it is likely prudent
SBP. Further, secondary to the vasodilated state leading to use prophylaxis in patients who have risk factors for
to decreased effective arterial circulating volume that is developing such infections.
characteristic of cirrhosis, albumin should be adminis-
Timing of Antibiotics in SBP and Septic Shock.
tered at diagnosis of SBP even without the obvious need
Recommendation. We suggest using appropriate
of volume resuscitation to prevent progression to ACLF.
antibiotics as soon as possible after recognition and
Typical initial dose is 1.5 g/kg of 25% albumin regardless
within 1 hour of shock onset in critically ill ACLF
of serum albumin levels.
patients with SBP and septic shock (Conditional rec-
Systemic Antifungal Prophylaxis for the Liver ommendation, low quality of evidence).
Transplant Recipient. Rationale. There are no RCTs to guide this recom-
Recommendation. We suggest using systemic antifungal mendation. The surviving sepsis guidelines recom-
prophylaxis in critically ill liver transplant recipients with mend initiating IV antibiotics as soon as possible after
risk factors for invasive fungal infections (Conditional recognition and within one hour for both sepsis and
recommendation, very low quality of evidence). septic shock. In an unselected patient population with
Recommendations. We suggest not using antifungal sepsis or septic shock, the timing and appropriateness
prophylaxis in critically ill liver transplant recipients of empiric antibiotic therapy demonstrated significant
at low risk for invasive fungal infections (Conditional impact on outcomes such as mortality, AKI, length of
recommendation, very low quality of evidence). stay, and acute lung injury.

664     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

Karvellas et al (38) in a retrospective cohort study n = 21). There were no statistically significant differ-
of SBP-associated septic shock from the Cooperative ences for mortality (OR, 1.42; 95% CI, 0.21–0.55),
Antimicrobial Therapy of Septic Shock database found renal impairment (OR, 3.00; 95% CI, 0.23–31.63)
that survivors compared with nonsurvivors were more or resolution of SBP (OR, 0.33; 95% CI, 0.03–3.51)
likely to receive appropriate antibiotic therapy as well (Supplementary Table 13, http://links.lww.com/
as receive therapy earlier. On multivariable adjustment, CCM/H302) (45).
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

each hour delay to appropriate antibiotic therapy was Selective Bowel Decontamination in the Liver
significantly associated with mortality (OR, 1.86; 95%
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

Transplant Candidate.
CI, 1.10–3.14 per hour increment). Similarly, Arabi et Recommendation. We suggest not using selective
al (39) from the same database found in a retrospective bowel decontamination (SBD) for critically ill liver
cohort of patients with cirrhosis and septic shock that transplant recipients (conditional recommendation,
the likelihood of death was significantly higher if ini- low quality of evidence).
tial therapy was either inappropriate (OR, 9.5; 95% CI, Rationale. Bacterial sepsis and wound complica-
4.3–20.7) or delayed (OR, 1.1; 95% CI, 1.1–1.2 for each tions after LT increase mortality, morbidity, or hos-
1 hr delay). Overall hospital mortality exceeded 75% in pital stay and are likely to increase overall transplant
both studies, which is significantly higher than other costs. All LT patients receive IV antibiotic prophylaxis
comparable septic shock studies. post-LT. The aim of SBD is to preemptively reduce
Both studies are at high risk of bias from their ret- aerobic Gram-negative bacterial and yeast carriage in
rospective nature and small sample sizes. The data the gut without elimination of anaerobic bacteria. A
from the general population are not directly applicable regimen typically consists of unabsorbed oral antibi-
to ACLF patients (Supplementary Table 12, http:// otics that have selective antimicrobial activity, with or
links.lww.com/CCM/H302). However, there is strong without a brief period of systemic antibiotic therapy.
rationale for the use of early appropriate antibiotic The use of SBD has not been widely adopted in North
therapy in SBP. This recommendation is applicable to America due to uncertainty regarding its net benefit
other infections in ACLF and ALF patients as well. to patients and potential it may promote the spread
Large Volume Paracentesis in SBP. of antibiotic resistance. Recently, Gurusamy et al (46)
Recommendation. We suggest not performing performed a systematic review of SBD of which iden-
large volume paracentesis (LVP) in critically ill ACLF tified four trials compared SBD versus placebo or no
patients with SBP (Conditional recommendation, very treatment (47–50). Including all four studies (n = 256
low quality of evidence). subjects), there were no statistically significant dif-
Remarks. LVP is defined as removing greater than ferences in rates of infection between patients who
4 L of ascitic fluid. received SBD and controls (RR, 0.94; 95% CI, 0.63–
Rationale. In patients with ACLF and ascites, SBP 1.41). In the three studies (n = 190 subjects) that re-
is a common complication and is associated with sig- ported mortality (47, 49, 50), there was no statistically
nificant mortality, particularly when co-existent with significant difference in mortality between patients
septic shock (38). As antibacterial activity in the as- who received SBD and controls (RR, 0.91; 95% CI,
citic fluid correlates with total protein, SBP occurs 0.31–2.72) (Supplementary Table 14, http://links.
commonly in patients with ascites of large volume and lww.com/CCM/H302). There were no significant dif-
low protein content (40, 41). LVP (defined as remov- ferences in pooled risk of graft rejection or retrans-
ing > 4 L of ascitic fluid) is widely used for the treat- plantation in reporting studies. Hence, given concerns
ment of refractory ascites. LVP may induce circulatory regarding potential side effects and risk of antibiotic
dysfunction, which can be mitigated with albumin as resistance, we cannot advocate for routine use of SBD
a plasma expanders (8 g/L ascites removed) (42–44). in ACLF/ALF patients undergoing LT.
However, there remains equipoise regarding the safety Initial Antibiotic Therapy for SBP.
and effectiveness of its use in patients with SBP. Choi Recommendation. We recommend using broad
et al (45) randomized 42 cirrhotic patients with SBP spectrum antibiotic agents for the initial management
to treatment with LVP (> 4L) and IV albumin (inter- of SBP in critically ill ACLF patients (Strong recom-
vention, n = 21) or diuretics and IV albumin (control, mendation, low quality of evidence).
Critical Care Medicine www.ccmjournal.org     665
Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

Rationale. SBP is a common life threatening com- (ESBL, MRSA, ± VRE) would be best suited for the
plication in cirrhosis (51). Delayed administration of empirical therapy. Once culture results are available,
appropriate antimicrobial therapy is associated with antibiotic therapy should be tailored to the narrowest
increased mortality (38, 39). Third-generation ceph- spectrum based on organism sensitivities.
alosporins are generally accepted agents of choice Midodrine and Terlipressin for SBP.
for empirical treatment of community acquired SBP Recommendation. We suggest not using mido-
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

(52). However, there is a trend of increased Gram- drine or terlipressin empirically for critically ill ACLF
positive and multidrug resistance pathogen, in-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

patients with SBP (Conditional recommendation, very

cluding methicillin-resistant Staphylococcus aureus low quality of evidence).
(MRSA), vancomycin-resistant enterococci (VRE), Rationale. SBP is a common infection in ACLF
and extended-spectrum beta-lactamase (ESBL) in patients. Patients with SBP are at increased risk of de-
multiple geographic areas that mandate careful con- veloping hepatorenal syndrome. Administration of al-
sideration of the initial treatment agent for SBP in set- bumin has been shown to reduce the risk of mortality
tings with high-drug resistance patterns (53, 54). Risk and renal impairment. Given the underlying vasodi-
factors associated with Gram-positive and multidrug- lated state, it is possible that administration of vaso-
resistant SBP are patients with advanced liver disease, pressors concomitant with albumin further reduces
severe critical illness, those receiving prophylactic the risk of renal injury and mortality. Salman et al (60)
antibiotics and nosocomial or community-acquired randomized 200 cirrhotic patients with SBP to one of
SBP (55, 56). A recent systematic review of the liter- four groups: albumin, terlipressin, low-dose albumin
ature (nine studies, 520 nosocomial SBP positive as- plus terlipressin, or midodrine. They failed to demon-
citic culture) revealed a remarkable high prevalence strate significant differences in mortality or renal im-
(30–66%) of multidrug-resistant pathogen, indicating pairment in any of the groups. On analysis of the data,
that third-generation cephalosporin may not be viable we found no differences in the risk of mortality (OR,
choices for nosocomial SBP (54). In another system- 1.63; 95% CI, 0.68–3.91), renal failure (OR, 2.63; 95%
atic review (seven studies, 1,701 participants), third- CI, 0.97–7.17), or resolution of SBP (OR, 0.48; 95%
generation cephalosporin-resistant pathogen were CI, 0.08–52.74) (Supplementary Table 16a, http://
reported in community (33.8%) as well as in nosoco- links.lww.com/CCM/H302) associated with the use
mial infections (54.3%), with pooled estimate indicat- of midodrine in SBP. Two RCTs informed our recom-
ing that nosocomial SBP was associated with a higher mendation regarding Terlipressin. In addition to the
risk for resistance compared with community acquired above study, Chelarescu et al (61) randomized 55 cir-
SBP (RR, 1.67; 95% CI, 1.14–2.44; p = 0.008) (57). For rhotic patients with SBP to cefotaxime or cefotaxime
healthcare-associated SBP, carbapenem-based empir- and terlipressin. The cefotaxime and terlipressin group
ical therapy was associated with lower rate of mor- had decreased mortality and increased resolution of
tality and treatment failure and more cost effectiveness SBP at 48 hours and 5 days as well as lower recurrence
compared with third-generation cephalosporin-based rates of SBP. On meta-analysis of the data from these
regimen (6% vs 25%; p = 0.01, 18% vs 51%; p = 0.001, two trials, we found no differences in the risk of mor-
respectively) (Supplementary Table 15, http://links. tality (OR, 0.66; 95% CI, 0.27–1.58), renal failure (OR,
lww.com/CCM/H302) (58, 59). Thus, we recommend 1.0; 95% CI, 0.32–3.09), or resolution of SBP (OR, 0.86;
limiting the use of third-generation cephalosporin as 95% CI, 0.67–5.15) associated with the use of terlip-
the initial empirical treatment to low-risk community- ressin in SBP (Supplementary Table 16b, http://links.
acquired SBP patients in the setting of low prevalence of lww.com/CCM/H302).
drug resistance. Active agents against ESBL-producing Both studies did not use albumin as standard of
pathogen (Carbapenems) should be considered for the care. Furthermore, both studies were at high risk of
empirical treatment of healthcare-associated SBP. In bias secondary to small sample sizes and lack of con-
high risk critically ill patients and nosocomial infec- cealment, blinding and description of randomization.
tions, tailored approach according to the antimicro- Further, the study by Chelarescu et al (61) was only
bial prevalence pattern covering resistant pathogens published in abstract form.

666     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

Gastroenterology Section reduces the risk of rebleeding rate but does not im-
pact mortality (73–75). We downgraded the level of
Timing of Endoscopy. evidence because across meta-analyses included stud-
Recommendation. We recommend performing ies were mostly retrospective and at high risk of bias
esophagogastroduodenoscopy no later than 12 hours from nonstandardized inclusion and treatment criteria
of presentation in critically ill ACLF patients with (Supplementary Table 17, http://links.lww.com/CCM/
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

portal hypertensive bleeding (known or suspected) H302). However, extrapolating from the indirect evi-
(Best Practice Statement). dence of the nonvariceal cohorts, short-term physio-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

Rationale. Acute portal hypertensive UGIB in a fre- logic benefits as well as the consistent demonstration of
quent complication in ACLF patients and often if the reduction in rebleeding across the studies, we issued a
triggering event for ACLF. The American Association strong recommendation in favor of PPIs.
of the Study of Liver Diseases recommends that endo-
scopic evaluation occur no later than 12 hours of pre- Octreotide or Somatostatin Analogs in Portal
sentation (62). There are no prospective data to guide Hypertensive Bleeding.
this recommendation. A recent meta-analysis compar- Recommendation. We recommend using octreo-
ing urgent (< 12 hr) versus nonurgent (> 12 hr) endos- tide or somatostatin analog (SSA) for the treatment
copy in acute variceal bleeding found that there were of portal hypertensive bleeding in critically ill patients
no differences in mortality, rebleeding rates, and other with ACLF (Strong recommendation, moderate quality
outcomes. However, this meta-analysis comprised five of evidence).
retrospective studies and was at high risk of selection Rationale. In patients with ACLF, acute variceal
bias (63). Given that early endoscopy would potentially bleeding is associated with mortality rates greater than
lead to earlier intervention and cessation of bleeding 10% per episode (76). Besides endoscopic variceal
source, reduce blood transfusions, and prevent hemo- banding or sclerotherapy, two classes of pharmacolog-
dynamic instability for continued bleeding, the panel ical agents for the treatment of acute variceal bleeding
strongly voted for early endoscopy. Because of the lack have been evaluated (77): terlipressin and its analogs
of high-quality data, we issued a BPS in favor of early (not available in North America) and SSAs (i.e., octreo-
endoscopy. tide). Based on pooled analysis of systematic reviews
of previous prospective controlled studies (78–80), the
Use of Proton Pump Inhibitors in Portal Hypertensive use of SSAs versus placebo was associated with 30 fewer
Bleeding. deaths per 1,000 patients (RR, 0.85; 95% CI, 0.72–1.00),
Recommendation. We recommend using proton although the effect on rebleeding outcome was less
pump inhibitors (PPIs) in critically ill ACLF patients clear (RR, 0.85; 95% CI, 0.52–1.37) (Supplementary
with portal hypertensive bleeding (Strong recommen- Table 18, http://links.lww.com/CCM/H302).
dation, low quality of evidence).
Rationale. PPIs block the final step of acid produc- Transjugular Intrahepatic Portosystemic Shunt for
tion by inhibiting hydrogen potassium ATPase in gas- Recurrent Variceal Bleeding.
tric parietal cells (64). In nonvariceal UGIB, they have Recommendation. We suggest using transjugular
consistently been shown to reduce rates of rebleeding, intrahepatic portosystemic shunt (TIPS) for recurrent
need for surgical or repeat endoscopic intervention (65). variceal bleeding after medical and endoscopic inter-
Potential mechanisms of benefit include stimulation of vention over continued endoscopic therapy in criti-
platelet aggregation and stabilization of fibrin clots by cally ill ACLF patients (conditional recommendation,
raising the gastric pH (66, 67). Whether these ben- low quality of evidence).
efits extend to portal hypertensive bleeding is unclear. Remark. TIPS requires appropriate screening for
Furthermore, the use of PPIs, especially in the popula- contraindications. This intervention requires access to
tion with cirrhosis is associated with alterations in the an experienced operator at a center with expertise.
microbiome leading to dysbiosis (68–70), increased risk Rationale. In patients with ACLF, the decision to
of SBP and hepatic encephalopathy (71) as well possibly prevent rebleeding after a significant variceal bleed is
increased mortality (72). Three meta-analyses found that a challenge. Traditionally, TIPS has been employed in
use of PPIs in patients with portal hypertensive bleeding the salvage/rescue setting after failure endoscopy. Most

Critical Care Medicine www.ccmjournal.org     667

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

recently, Garcia-Pagan et al (81) demonstrated in a There are no randomized trials to guide recom-
randomized trial of 63 cirrhotic/ACLF patients at high mendations. In heterogeneous critically ill patients,
risk of treatment failure that patients who underwent relief of intra-abdominal hypertension is associated
TIPS within 72 hours post-bleed after randomization with improvements in organ function and outcomes
that rebleeding rates (3% vs 50%) and mortality (14% (88, 89). In ACLF patients with tense ascites and raised
vs 39%; p < 0.001 for both) were significant lower in intra-abdominal pressure, drainage of ascites lowers
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

the early TIPS group compared with pharmacotherapy/ intra-abdominal pressure. There is a strong physiologic
band ligation. In a recent meta-analysis, Halabi et al (82) rationale to attempt a trial of LVP in ACLF patients, es-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

demonstrated that in nine RCTs involving 608 cirrhotic pecially if concomitant intra-abdominal hypertension
patients, early TIPS was associated with decreased is present. Therefore, we issued a BPS in favor of LVP.
1-year mortality (RR, 0.68; 95% CI, 0.49–0.96; p = 0.03)
and 1-year occurrence rate of rebleeding (RR, 0.28; Peri-Transplant Section
95% CI, 0.20–0.40; p < 0.001). No significant difference
in the occurrence rate of hepatic encephalopathy at 1 Corticosteroid Administration to Deceased Donors.
year was observed (RR, 1.36; 95% CI, 0.72–2.56; p = Recommendation. We suggest using systemic corti-
0.34). While our systematic review of 11 studies did not costeroids for deceased liver graft donors (Conditional
demonstrate increased rates of hepatic encephalopathy recommendation, very low quality of evidence).
(RR, 1.36; 95% CI, 0.72–2.56) (Supplementary Table Rationale. In a systematic review of brain-dead
19, http://links.lww.com/CCM/H302), in patients with organ donors (of any organ) (90), the pooled results
a model for end-stage liver disease (MELD) greater of RCTs (91, 92) demonstrated that liver grafts from
than 20 or significant hepatic encephalopathy, consider 183 deceased donors receiving corticosteroids showed
the use of TIPS on a case-by-case basis. a reduction in post-transplantation graft dysfunction
(4.2% absolute risk reduction; 91 fewer to 72 more
LVP in Intra-Abdominal Hypertension.
per 1,000; Supplementary Table 20, http://links.lww.
Recommendation. We recommend performing LVP
com/CCM/H302) compared with grafts from the con-
with measurement of intra-abdominal pressure in
trol group. Please refer to the SCCM “Guidelines for
critically ill ACLF patients with tense ascites and intra-
the Management of the Potential Organ Donor in the
abdominal hypertension or hemodynamic, renal or
ICU” (93).
respiratory compromise (Best Practice Statement).
Rationale. Ascites is a common complication in Fluid Management of Deceased Donor.
patients with ACLF. When ascites becomes tense, Recommendation. We suggest either using goal-
renal respiratory and cardiovascular function maybe directed fluid management for the deceased organ
compromised from rises in intra-abdominal pres- donor or standard fluid management strategies
sure (83–86). Secondary to the vasodilated state of (Conditional recommendation, very low quality of
liver disease and limited compensatory mechanisms, evidence).
critical abdominal organ hypoperfusion may occur Remarks. Goal-directed fluid management refers
in ACLF patients with tense ascites. In a study of 22 to management directed by invasive hemodynamic
critically ill patients with decompensated cirrhosis and monitoring (measurement of filling pressures, car-
intra-abdominal hypertension, Mayr et al (87) dem- diac output, and central venous oximetry). In contrast,
onstrated reduced clearance of indo-cyanine green standard fluid management refers to management
(ICG) dye which dramatically improved upon LVP. based on clinical assessment of peripheral perfusion
Concomitantly hepatic artery resistance and blood (e.g., capillary refill time).
flow velocities improved, intra-abdominal pressure Rationale. Goal-directed fluid management of de-
fell, and abdominal perfusion pressure rose. Mayr et al ceased donors, compared with standard management,
(87) attributed the ICG clearance changes to improved is associated with negligible desirable effects (1 per
hepatosplanchnic blood flow. Observational studies 1,000 absolute reduction in mortality, range 24 fewer
have also demonstrated improvement in lung func- to 33 more deaths) (Supplementary Table 21, http://
tion and Pao2/Fio2 ratio upon LVP concomitant with links.lww.com/CCM/H302), and a small likelihood
decreases in intra-abdominal pressure (86). of undesirable effects due to delays or complications

668     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

related to invasive monitoring and fluid overload Remarks. Providers may choose to use artificial liver
(94). No data directly addressed the impact of goal- support based on local availability, familiarity with its
directed fluid management or other components of use, and available resources.
goal-directed donor management specifically on the Rationale. Extracorporeal liver support is used as a
outcomes of liver recipients. The SCCM “Guidelines bridge to transplant or spontaneous recovery in ALF
for the Management of the Potential Organ Donor in and as a bridge to transplant in ACLF. Based upon
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

the ICU” recommends maintaining euvolemia in the pooled data from 24 RCTs (n = 1,778), which included
donor (mean arterial pressure at least 60 mm Hg, urine patients with either ALF or ACLF, the desirable effects
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

output of 1 mL/kg/hr, left ventricle ejection fraction of liver support (artificial and bioartificial combined)
> 45%) using an isotonic crystalloid and low doses of range from 3.1% absolute reduction in mortality for
vasopressor (e.g., ≤ 10 µg/kg/min); pulmonary artery acute liver disease (range for acute liver disease: 85
or central venous catheter or noninvasive monitoring fewer to 36 more deaths per 1,000) to 11.5% absolute
should be considered to guide fluid management (93). reduction for acute-on-chronic liver disease (range for
acute-on-chronic liver disease: 180 fewer to 42 more
Recipient Acuity and Donor Assessment.
deaths per 1,000); neither mortality reduction is sta-
Recommendation. There was insufficient evidence to
tistically significant (Supplementary Table 22, http://
issue a recommendation on using the donor risk index
links.lww.com/CCM/H302). The selection of bioarti-
(DRI) in selection of liver allograft.
ficial support systems is further limited by feasibility
Remark. Clinicians should use their judgment re- (98). Artificial liver support has small desirable effects,
garding severity of illness of the potential transplant moderate undesirable effects and is associated with
recipient with donor graft factors (i.e., cold ischemia high costs and limited access.
time, steatosis, donor age, etc).
Rationale. Based upon low-quality evidence from Peri-Transplant Fluid Restriction Accompanied
three observational studies that were unable to be by Vasopressor Support in the Liver Transplant
pooled, the DRI of the graft did not appear to affect Recipient.
patient survival. Two of the three studies found graft Recommendation. There was insufficient evidence
factors were associated with graft survival. One study to issue a recommendation on peri-transplant fluid
(n = 1,090) found that a high DRI graft (> 1.8) may restriction accompanied by vasopressor use in liver
adversely affect graft survival, particularly in recipi- transplant recipients.
ents with low and intermediate MELD scores; how- Rationale. We were unable to identify high-quality
evidence addressing whether low central venous pres-
ever, in recipients with high MELD scores (> 30), graft
sure (CVP) and vasopressor infusion impacts patient or
survival appeared to be similar for low and high DRI
graft survival in LT. A 2011 Cochrane review addressed
grafts (95). A second, smaller study (n = 115) used
the impact of low CVP and vasopressor use; however,
three categories of graft risk (standard graft, 1–2 risk
mortality and graft survival were not reported (99).
factors, 3–4 risk factors). Graft risk factors were asso-
Mean blood transfusion volume was reduced by low
ciated with graft, but not patient, survival (96). The
CVP (1.2 L lower; range: 1.63 lower to 0.77 lower) com-
remaining observational trial (n = 70) compared two
pared with controls. There were no significant differ-
categories of grafts (more than one extended donor
ences in peri-transplant renal function or postoperative
criteria [EDC] vs grafts with none or one EDC) and
complications in the low CVP group. Norepinephrine
found no difference in early (5-d post-transplant)
use, compared with control, resulted in no significant
graft function (97).
difference in allogeneic blood transfusion require-
Extracorporeal Liver Support for Acute or Acute-on- ments, platelets volume transfused, or plasma volume
Chronic Liver Failure. transfused. An increasingly common intraoperative
Recommendation. We suggest using either extracor- practice is to restrict fluid during the preanhepatic
poreal liver support or standard medical therapy in and anhepatic stages in the liver transplant recipient in
critically ill ALF or ACLF patients (Conditional rec- order to lessen transfusion requirements. Mean arterial
ommendation, very low quality of evidence). pressure may be supported by vasopressors as needed.

Critical Care Medicine www.ccmjournal.org     669

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

TABLE 2.
Summary of Recommendations
Strength of Quality of
Recommendation Recommendation Evidence
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

We recommend performing esophagogastroduodenoscopy no later than 12 hr Best practice statement Best practice
of presentation in critically ill ACLF patients with portal hypertensive bleeding statement
(known or suspected)
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

We recommend performing LVP with measurement of intra-abdominal pressure in Best practice statement Best practice
critically ill ACLF patients with tense ascites and intra-abdominal hypertension statement
or hemodynamic, renal or respiratory compromise
We recommend using antibiotic prophylaxis in critically ill ACLF patients with any Strong Moderate
type of upper gastrointestinal bleeding
We recommend using albumin in critically ill ACLF patients with SBP Strong Moderate
We recommend using octreotide or somatostatin analog for the treatment of Strong Moderate
portal hypertensive bleeding in critically ill ACLF patients
We recommend using proton pump inhibitors in critically ill ACLF patients with Strong Low
portal hypertensive bleeding
We recommend using broad spectrum antibiotic agents for the initial management Strong Low
of SBP in critically ill ACLF patients
We suggest, when available, using plasma exchange in critically ill ALF patients Conditional Low
who develop hyperammonemia
We suggest using hypertonic saline in critically ill ALF patients who are at risk of Conditional Low
developing intracranial hypertension
We suggest using nonabsorbable disaccharide in critically ill ACLF patients with Conditional Low
overt hepatic encephalopathy
We suggest using enteral polyethylene glycol as an alternative to lactulose in criti- Conditional Low
cally ill ACLF with overt hepatic encephalopathy
We suggest using oral rifaximin as adjunctive therapy in critically ill ACLF patients Conditional Low
with overt hepatic encephalopathy
We suggest using appropriate antibiotics as soon as possible after recognition Conditional Low
and within 1 hr of shock onset in critically ill ACLF patients with SBP and septic
shock
We suggest not using selective bowel decontamination for the critically ill liver Conditional Low
transplant recipient
We suggest using transjugular intrahepatic portosystemic shunt in critically ill Conditional Low
ACLF patients with recurrent variceal bleeding after medical and endoscopic
intervention over continued endoscopic therapy
We suggest using balanced (or normochloremic) crystalloid solution over normal Conditional Low
(hyperchloremic) saline for peri-transplant fluid replacement in liver transplant
recipients
We suggest using albumin over crystalloid for intraoperative volume replacement Conditional Low
during liver transplantation
We suggest not using invasive intracranial pressure monitoring for critically ill ALF Conditional Very low
patients with advanced-grade encephalopathy
We suggest not routinely using induced moderate hypothermia (< 34°C) for criti- Conditional Very low
cally ill ALF patients who are at risk of developing intracranial hypertension
We suggest using LOLA in critically ill ACLF patients with overt hepatic Conditional Very low
encephalopathy
(Continued)

670     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

TABLE 2. (Continued).
Summary of Recommendations
Strength of Quality of
Recommendation Recommendation Evidence
We suggest not routinely using IV flumazenil, zinc supplementation, glycerol phen- Conditional Very low
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

ylbutyrate, probiotics, or acarbose as adjunctive therapies in critically ill patients

ACLF patients with overt hepatic encephalopathy
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

We suggest using systemic antifungal prophylaxis in critically ill liver transplant Conditional Very low
recipients with risk factors for invasive fungal infections
We suggest not using antifungal prophylaxis in critically ill liver transplant recipi- Conditional Very low
ents at low risk for invasive fungal infections
We suggest not performing LVP in critically ill ACLF patients with SBP Conditional Very low
We suggest not using midodrine or terlipressin for critically ill ACLF patients with Conditional Very low
SBP
We suggest using systemic corticosteroids for deceased liver graft donors Conditional Very low
We suggest either using goal-directed fluid management for the deceased organ Conditional Very low
donor or standard fluid management strategies
We suggest using either extracorporeal liver support or standard medical therapy Conditional Very low
in critically ill ALF or ACLF patients
There was insufficient evidence to issue a recommendation on using lactulose, Not applicable Not applicable
rifaximin, flumazenil, branch-chain amino acids, carnitine, zinc, probiotics, and
LOLA in critically ill ALF patients with hyperammonemia
There was insufficient evidence to issue a recommendation on using the donor Not applicable Not applicable
risk index in selection of liver allograft
There was insufficient evidence to issue a recommendation on peri-transplant fluid Not applicable Not applicable
restriction accompanied by vasopressor use in liver transplant recipients
There was insufficient evidence to issue a recommendation for the choice of intra- Not applicable Not applicable
operative monitoring in liver transplantation recipients
There was insufficient evidence to issue recommendation on early extubation of Not applicable Not applicable
liver transplant recipients
ACLF = acute on chronic liver failure, ALF = acute liver failure, LOLA = L-ornithine L-aspartate, LVP = large volume paracentesis,
SBP = spontaneous bacterial peritonitis.

Fluid Management: Choice of Peri-Transplant Crystalloid. and found no mortality difference; however, the
Recommendation. We suggest using balanced (or total number of deaths was low. There was no differ-
normochloremic) crystalloid solution over normal ence in need for renal replacement therapy between
(hyperchloremic) saline for peri-transplant fluid re- groups (101). A recently published RCT of 7,900 crit-
placement in liver transplant recipients (Conditional ically ill patients from five ICUs showed an absolute
recommendation, low quality of evidence). reduction in adjusted mortality of 20 patients per
Rationale. We found no direct evidence comparing 1,000 (range: from 12 more to 45 fewer per 1,000) in
different types of crystalloids and risk of survival or the normochloremic (balanced) crystalloids group.
graft failure after LT. In a 2014 meta-analysis, indirect Major adverse kidney events were also reduced in the
evidence (in nonliver transplant populations) showed
normochloremic group (Supplementary Table 23,
that balanced crystalloid, compared with normal sa-
http://links.lww.com/CCM/H302) (102).
line, improved survival in sepsis patients (low-level
evidence) (100). A 2017 Cochrane review of sur- Fluid Management: Crystalloid Versus Colloids.
gical patients evaluated 18 RCTs of 1,096 patients Recommendation. We suggest using albumin over crys-
receiving either buffered (normochloremic or bal- talloid for intraoperative volume replacement during LT
anced) or nonbuffered (normal saline) crystalloid (Conditional recommendation, low quality of evidence).

Critical Care Medicine www.ccmjournal.org     671

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

Remark. Starches should not be used due to the risk (neurology, infectious disease, gastroenterology, and
of coagulopathy and renal failure. peri-transplant). A summary list of recommenda-
Rationale. For patients undergoing LT, no studies tions is provided in Table 2. We assembled multidis-
were identified comparing the effects of colloids versus ciplinary experts to address pertinent questions that
crystalloids on mortality or graft survival. Using indi- are commonly encountered by clinicians taking care
rect evidence (patients with traumatic injuries, patients of patients with ALF and ACLF. We used a rigorous
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

undergoing surgery and critically ill patients), a meta- methodological approach lead by international experts
analysis showed decrease mortality with albumin (ab- in methodology to summarize the evidence and subse-
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

solute mortality 47 fewer patients per 1,000; range: 95 quently used the expertise of content experts to issue
fewer to seven more deaths per 1,000) (Supplementary recommendations. Our approach led to the generation
Table 24, http://links.lww.com/CCM/H302) (100). In of a contemporary document that can be used as a ref-
another meta-analysis, colloid administration with erence for clinicians. There are some important limita-
starch (tetrastarch, pentastarch, dextran, and gelatin) tions of this guideline, which include the lack of patient
increased the risk of renal replacement therapy without participation in the guideline development process,
a difference in morality (103). although panel members focused on the patient per-
Intraoperative Hemodynamic Monitoring. spective when issuing the recommendations; it is pos-
Recommendation. There was insufficient evidence to sible that this perspective does not entirely reflect the
issue a recommendation for the choice of intraopera- values and preferences of patients. Last, we were un-
tive monitoring in LT recipients. able to comment on other pertinent PICO questions
Rationale. Many studies have compared traditional that were not prioritized by the guideline committee.
hemodynamic monitors to newer monitoring tech- However, we identified several areas where evidence
niques primarily in terms of measurement accuracy for this population is lacking and should be targeted
and other performance characteristics; however, no for future research.
studies of newer monitors (including transesophageal
echocardiography) were designed to show improve- 1 Division of Pulmonary and Critical Care Medicine, Medical
College of Wisconsin, Milwaukee, WI.
ments in patient or graft survival.
2 Emory University Hospital, Atlanta, GA.
Early Extubation of Liver Transplant Recipients. 3 Department of Medicine, McMaster University, Hamilton,
Recommendation. There was insufficient evidence ON, Canada.
to issue recommendation on early extubation of liver 4 Froedtert and the Medical College of Wisconsin, Milwaukee,
transplant recipient. WI.
Remark. Clinicians should use clinical judgment 5 Mayo Clinic, Phoenix, AZ.
based on center expertise and recipient status. 6 Medical College of Wisconsin, Milwaukee, WI.
Rationale. New evidence is emerging regarding 7 University of Pittsburgh Medical Center, Pittsburgh, PA.
decreased respiratory complications with early extu- 8 GUIDE Group, McMaster University, Hamilton, ON, Canada.
bation post-LT. Among liver transplant recipients, 9 Mayo Clinic, Rochester, MN.
patients who received anesthetic technique using 10 Hackensack University Medical Center, Hackensack, NJ.
shorter acting agents (vs traditional anesthetic tech- 11 Henry Ford Health System, Detroit, MI.
nique) were extubated sooner (553 vs 1,081 min; p < 12 University of Michigan Hospitals, Ann Arbor, MI.
0.001) but spent similar duration in the ICU (104). The 13 Northeast Georgia Medical Center, Gainesville, GA.
study was not designed to assess patient’s or graft sur- 14 Health Research Methods, Evidence, and Impact, McMaster
University, Hamilton, ON, Canada.
vival. Institutional staffing and ICU service environ-
15 University of Kentucky College of Medicine, Lexington, KY.
ments appear to affect post-transplant disposition and
16 Kansas University Medical Center, Kansas City, KS.
length of post-transplant ventilation.
17 University of California Los Angeles Medical Center, Los
Angeles, CA.
DISCUSSION 18 Barnes Jewish Hospital, St. Louis, MO.
19 Department of Critical Care Medicine and Division of
We report 29 recommendations on the manage-
Gastroenterology (Liver Unit), University of Alberta,
ment ALF or ACLF in the ICU, related to four groups Edmonton, AB, Canada.

672     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

Supplemental digital content is available for this article. Direct ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed
URL citations appear in the printed text and are provided in the August 23, 2022
HTML and PDF versions of this article on the journal’s website 6. Guyatt GH, Oxman AD, Vist GE, et al; GRADE Working Group:
(http://journals.lww.com/ccmjournal). GRADE: An emerging consensus on rating quality of evidence
Conflicts of interest were reviewed and adjudicated by the co- and strength of recommendations. BMJ 2008; 336:924–926
chairs and co-vice chairs of the guidelines. In the event an indi- 7. GRADEpro GDT: GRADEpro Guideline Development Tool
vidual disclosed a conflict or potential conflict by submitted form [Software]. 2015. Available at: gradepro.org. Accessed
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

or verbally during the process of guidelines, those individuals February 15, 2022
abstained from voting on related questions. The taskforce fol- 8. Neumann I, Brignardello-Petersen R, Wiercioch W, et al: The
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

lowed all procedures as documented in the American College of GRADE evidence-to-decision framework: A report of its
Critical Care Medicine/Society of Critical Care Medicine (SCCM) testing and application in 15 international guideline panels.
Standard Operating Procedures Manual. Drs. Singbartl, Nanchal, Implement Sci 2016; 11:93
Killian, Olson, Karvellas, Subramanian, and Truwit disclosed au- 9. Alexander PE, Gionfriddo MR, Li SA, et al: A number of fac-
thorship on several related articles with potential intellectual con- tors explain why WHO guideline developers make strong
flicts explored and adjudicated. Dr. Dionne described volunteer recommendations inconsistent with GRADE guidance. J Clin
service for Canadian Association of Gastroenterology, American Epidemiol 2016; 70:111–122
College of Gastroenterology, American Gastroenterological
10. Guyatt GH, Schunemann HJ, Djulbegovic B, et al: Guideline
Association, and European Society of Intensive Care Medicine. Dr.
panels should not GRADE good practice statements. J Clin
Hyzy described volunteer service for American Thoracic Society,
Epidemiol 2015; 68:597–600
Quality Improvement and Implementation Committee, and the
SCCM Finance Committee as well as service as an expert witness 11. Karvellas CJ, Fix OK, Battenhouse H, et al; U S Acute Liver
in a previous medical case involving this subject matter. Dr. Taylor Failure Study Group: Outcomes and complications of intracra-
advised of service as an author on the SCCM/American Society nial pressure monitoring in acute liver failure: A retrospective
of Parenteral and Enteral Nutrition (ASPEN) nutrition guidelines cohort study. Crit Care Med 2014; 42:1157–1167
and service on the ASPEN research committee. Dr. Huang dis- 12. Keays RT, Alexander GJ, Williams R: The safety and value of
closed service on the American College of Emergency Physicians extradural intracranial pressure monitors in fulminant hepatic
Sepsis Task Force. Dr. Karvellas disclosed service on an Acute failure. J Hepatol 1993; 18:205–209
Liver Failure Study Group. Dr. Hollenberg disclosed that he is 13. Daas M, Plevak DJ, Wijdicks EF, et al: Acute liver failure:
a member of American College of Chest Physicians, American Results of a 5-year clinical protocol. Liver Transpl Surg 1995;
Heart Association, and American College of Cardiology. Dr. 1:210–219
Olson disclosed that she is a member of American Association 14. Gluud LL, Vilstrup H, Morgan MY: Nonabsorbable disaccha-
for the Study of Liver Disease. Dr. Steadman disclosed that he rides for hepatic encephalopathy: A systematic review and
is a co-author on “Management of the Critically Ill Patient with meta-analysis. Hepatology 2016; 64:908–922
Cirrhosis: A Multidisciplinary Perspective.” J Hepatology 2015 15. Kimer N, Krag A, Moller S, et al: Systematic review with meta-
pending publication, and that he is currently writing guidelines analysis: The effects of rifaximin in hepatic encephalopathy.
for Anesthesiology Transplant fellowship for the International Liver Aliment Pharmacol Ther 2014; 40:123–132
Transplant Society. The remaining authors have disclosed that 16. Dalal R, McGee RG, Riordan SM, et al: Probiotics for people
they do not have any potential conflicts of interest. with hepatic encephalopathy. Cochrane Database Syst Rev
For information regarding this article, E-mail: rnanchal@mcw.edu 2017; 2:CD008716
17. Gluud LL, Dam G, Les I, et al: Branched-chain amino acids for
people with hepatic encephalopathy. Cochrane Database Syst
Rev 2017; 5:CD001939
REFERENCES
18. Goh ET, Andersen ML, Morgan MY, et al: Flumazenil versus
1. Nanchal R, Subramanian R, Karvellas CJ, et al: Guidelines for placebo or no intervention for people with cirrhosis and he-
the management of adult acute and acute-on-chronic liver patic encephalopathy. Cochrane Database Syst Rev 2017;
failure in the ICU: Cardiovascular, endocrine, hematologic, 8:CD002798
pulmonary, and renal considerations. Crit Care Med 2020; 19. Goh ET, Stokes CS, Sidhu SS, et al: L-ornithine L-aspartate
48:e173–e191 for prevention and treatment of hepatic encephalopathy in
2. Moreau R, Jalan R, Gines P, et al: Acute-on-chronic liver people with cirrhosis. Cochrane Database Syst Rev 2018;
failure is a distinct syndrome that develops in patients with 5:CD012410
acute decompensation of cirrhosis. Gastroenterology 2013; 20. Shores NJ, Keeffe EB: Is oral L-acyl-carnitine an effective
144:1426–1437, 1437.e1–e9 therapy for hepatic encephalopathy? Review of the literature.
3. O’Grady JG, Williams R: Classification of acute liver failure. Dig Dis Sci 2008; 53:2330–2333
Lancet 1993; 342:743 21. Bernal W, Hall C, Karvellas CJ, et al: Arterial ammonia
4. Higgins JP, Altman DG, Gotzsche PC, et al: The Cochrane and clinical risk factors for encephalopathy and intracra-
Collaboration’s tool for assessing risk of bias in randomised nial hypertension in acute liver failure. Hepatology 2007;
trials. BMJ 2011; 343:d5928 46:1844–1852
5. Wells GA, Shea B, O'Connell D, et al: The Newcastle-Ottawa 22. Larsen FS, Schmidt LE, Bernsmeier C, et al: High-volume
Scale (NOS) for Assessing the Quality If Nonrandomized plasma exchange in patients with acute liver failure: An open
Studies in Meta-Analyses. 2009. Available at: https://www. randomised controlled trial. J Hepatol 2016; 64:69–78

Critical Care Medicine www.ccmjournal.org     673

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

23. Murphy N, Auzinger G, Bernel W, et al: The effect of hy- peritonitis-associated septic shock: A retrospective cohort
pertonic sodium chloride on intracranial pressure in study. Aliment Pharmacol Ther 2015; 41:747–757
patients with acute liver failure. Hepatology 2004; 39: 39. Arabi YM, Dara SI, Memish Z, et al; Cooperative Antimicrobial
464–470 Therapy of Septic Shock (CATSS) Database Research Group:
24. Jalan R, SW OD, Deutz NE, et al: Moderate hypothermia for Antimicrobial therapeutic determinants of outcomes from
uncontrolled intracranial hypertension in acute liver failure. septic shock among patients with cirrhosis. Hepatology 2012;
Lancet 1999; 354:1164–1168 56:2305–2315
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

25. Jalan R, Olde Damink SW, Deutz NE, et al: Moderate hy- 40. Runyon BA: Low-protein-concentration ascitic fluid is predis-
pothermia in patients with acute liver failure and uncon- posed to spontaneous bacterial peritonitis. Gastroenterology
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

trolled intracranial hypertension. Gastroenterology 2004; 1986; 91:1343–1346

127:1338–1346 41. Andreu M, Sola R, Sitges-Serra A, et al: Risk factors for spon-
26. Vaquero J: Therapeutic hypothermia in the management of taneous bacterial peritonitis in cirrhotic patients with ascites.
acute liver failure. Neurochem Int 2012; 60:723–735 Gastroenterology 1993; 104:1133–1138
27. Larsen FS, Murphy N, Bernal W, et al; EUROALF Group: The 42. Sola R, Andreu M, Coll S, et al: Spontaneous bacterial per-
prophylactive effect of mild hypothermia to prevent brain itonitis in cirrhotic patients treated using paracentesis or
edema in patients with acute liver failure: Results of a multi- diuretics: Results of a randomized study. Hepatology 1995;
center randomized, controlled trial (abstract). J Hepatol 2011; 21:340–344
54(Suppl 1):S26 43. Arora V, Vijayaraghavan R, Maiwall R, et al: Paracentesis-
28. Karvellas CJ, Todd Stravitz R, Battenhouse H, et al: Therapeutic induced circulatory dysfunction with modest-volume
hypothermia in acute liver failure: A multicenter retrospective paracentesis is partly ameliorated by albumin infusion in acute-
cohort analysis. Liver Transpl 2015; 21:4–12 on-chronic liver failure. Hepatology 2020; 72:1043–1055
29. Bernal W, Murphy N, Brown S, et al: A multicentre random- 44. Knauer CM, Lowe HM: Hemodynamics in the cirrhotic patient
ized controlled trial of moderate hypothermia to prevent in- during paracentesis. N Engl J Med 1967; 276:491–496
tracranial hypertension in acute liver failure. J Hepatol 2016; 45. Choi CH, Ahn SH, Kim DY, et al: Long-term clinical outcome of
65:273–279 large volume paracentesis with intravenous albumin in patients
30. Rahimi RS, Singal AG, Cuthbert JA, et al: Lactulose vs polyeth- with spontaneous bacterial peritonitis: A randomized prospec-
ylene glycol 3350--electrolyte solution for treatment of overt tive study. J Gastroenterol Hepatol 2005; 20:1215–1222
hepatic encephalopathy: The HELP randomized clinical trial. 46. Gurusamy KS, Nagendran M, Davidson BR: Methods of
JAMA Intern Med 2014; 174:1727–1733 preventing bacterial sepsis and wound complications after
31. Chavez-Tapia NC, Cesar-Arce A, Barrientos-Gutierrez T, et liver transplantation. Cochrane Database Syst Rev 2014;
al: A systematic review and meta-analysis of the use of oral (3):CD006660
zinc in the treatment of hepatic encephalopathy. Nutr J 2013; 47. Bion JF, Badger I, Crosby HA, et al: Selective decontamination
12:74 of the digestive tract reduces gram-negative pulmonary colo-
32. Rockey DC, Vierling JM, Mantry P, et al; HALT-HE Study nization but not systemic endotoxemia in patients undergoing
Group: Randomized, double-blind, controlled study of glycerol elective liver transplantation. Crit Care Med 1994; 22:40–49
phenylbutyrate in hepatic encephalopathy. Hepatology 2014; 48. Arnow PM, Carandang GC, Zabner R, et al: Randomized con-
59:1073–1083 trolled trial of selective bowel decontamination for preven-
33. Gentile S, Guarino G, Romano M, et al: A randomized controlled tion of infections following liver transplantation. Clin Infect Dis
trial of acarbose in hepatic encephalopathy. Clin Gastroenterol 1996; 22:997–1003
Hepatol 2005; 3:184–191 49. Hellinger WC, Yao JD, Alvarez S, et al: A randomized, pro-
34. Chavez-Tapia NC, Barrientos-Gutierrez T, Tellez-Avila F, et al: spective, double-blinded evaluation of selective bowel de-
Meta-analysis: Antibiotic prophylaxis for cirrhotic patients with contamination in liver transplantation. Transplantation 2002;
upper gastrointestinal bleeding - an updated Cochrane review. 73:1904–1909
Aliment Pharmacol Ther 2011; 34:509–518 50. Zwaveling JH, Maring JK, Klompmaker IJ, et al: Selective de-
35. Salerno F, Navickis RJ, Wilkes MM: Albumin infusion improves contamination of the digestive tract to prevent postoperative
outcomes of patients with spontaneous bacterial peritonitis: A infection: A randomized placebo-controlled trial in liver trans-
meta-analysis of randomized trials. Clin Gastroenterol Hepatol plant patients. Crit Care Med 2002; 30:1204–1209
2013; 11:123–130.e1 51. Tandon P, Garcia-Tsao G: Bacterial infections, sepsis, and mul-
36. Eschenauer GA, Kwak EJ, Humar A, et al: Targeted versus uni- tiorgan failure in cirrhosis. Semin Liver Dis 2008; 28:26–42
versal antifungal prophylaxis among liver transplant recipients. 52. Jalan R, Fernandez J, Wiest R, et al: Bacterial infections in
Am J Transplant 2015; 15:180–189 cirrhosis: A position statement based on the EASL Special
37. Evans JD, Morris PJ, Knight SR: Antifungal prophylaxis in liver Conference 2013. J Hepatol 2014; 60:1310–1324
transplantation: A systematic review and network meta-analy- 53. Friedrich K, Nussle S, Rehlen T, et al: Microbiology and re-
sis. Am J Transplant 2014; 14:2765–2776 sistance in first episodes of spontaneous bacterial peritonitis:
38. Karvellas CJ, Abraldes JG, Arabi YM, et al; Cooperative Implications for management and prognosis. J Gastroenterol
Antimicrobial Therapy of Septic Shock (CATSS) Database Hepatol 2016; 31:1191–1195
Research Group: Appropriate and timely antimicrobial 54. Fiore M, Maraolo AE, Gentile I, et al: Nosocomial sponta-
therapy in cirrhotic patients with spontaneous bacterial neous bacterial peritonitis antibiotic treatment in the era of

674     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Special Articles

multi-drug resistance pathogens: A systematic review. World J 70. Brandl K, Schnabl B: Is intestinal inflammation linking dysbio-
Gastroenterol 2017; 23:4654–4660 sis to gut barrier dysfunction during liver disease? Expert Rev
55. Alexopoulou A, Papadopoulos N, Eliopoulos DG, et al: Gastroenterol Hepatol 2015; 9:1069–1076
Increasing frequency of gram-positive cocci and gram-nega- 71. Dam G, Vilstrup H, Watson H, et al: Proton pump inhibitors
tive multidrug-resistant bacteria in spontaneous bacterial per- as a risk factor for hepatic encephalopathy and spontaneous
itonitis. Liver Int 2013; 33:975–981 bacterial peritonitis in patients with cirrhosis with ascites.
56. Kim JH, Jeon YD, Jung IY, et al: Predictive factors of sponta- Hepatology 2016; 64:1265–1272
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

neous bacterial peritonitis caused by gram-positive bacteria in 72. Dultz G, Piiper A, Zeuzem S, et al: Proton pump inhibitor
patients with cirrhosis. Medicine (Baltim) 2016; 95:e3489 treatment is associated with the severity of liver disease and
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

57. Fiore M, Gentile I, Maraolo AE, et al: Are third-generation increased mortality in patients with cirrhosis. Aliment Pharmacol
cephalosporins still the empirical antibiotic treatment of com- Ther 2015; 41:459–466
munity-acquired spontaneous bacterial peritonitis? A system- 73. Lin L, Cui B, Deng Y, et al: The efficacy of proton pump inhib-
atic review and meta-analysis. Eur J Gastroenterol Hepatol itor in cirrhotics with variceal bleeding: A systemic review and
2018; 30:329–336 meta-analysis. Digestion 2021; 102:117–127
58. Merli M, Lucidi C, Di Gregorio V, et al: An empirical broad spec- 74. Zhu J, Qi X, Yu H, et al: Acid suppression in patients treated
trum antibiotic therapy in health-care-associated infections with endoscopic therapy for the management of gastroesoph-
improves survival in patients with cirrhosis: A randomized trial. ageal varices: A systematic review and meta-analysis. Expert
Hepatology 2016; 63:1632–1639 Rev Gastroenterol Hepatol 2018; 12:617–624
59. Lucidi C, Di Gregorio V, Ceccarelli G, et al: A cost analysis 75. Lo EA, Wilby KJ, Ensom MH: Use of proton pump inhibitors in
of a broad-spectrum antibiotic therapy in the empirical treat- the management of gastroesophageal varices: A systematic
ment of health care-associated infections in cirrhotic patients. review. Ann Pharmacother 2015; 49:207–219
Clinicoecon Outcomes Res 2017; 9:385–390 76. Carbonell N, Pauwels A, Serfaty L, et al: Improved survival after
60. Salman TA, Edrees AM, El-Said HH, et al: Effect of different variceal bleeding in patients with cirrhosis over the past two
therapeutic modalities on systemic, renal, and hepatic hemo- decades. Hepatology 2004; 40:652–659
dynamics and short-term outcomes in cirrhotic patients with 77. Escorsell A, Ruiz del Arbol L, Planas R, et al: Multicenter ran-
spontaneous bacterial peritonitis. Eur J Gastroenterol Hepatol domized controlled trial of terlipressin versus sclerotherapy
2016; 28:777–785 in the treatment of acute variceal bleeding: The TEST study.
61. Chelarescu O, Chelarescu D, Stratan I, et al: Terlipressin influ- Hepatology 2000; 32:471–476
ence in spontaneous bacterial peritonitis. EASL International 78. Wells M, Chande N, Adams P, et al: Meta-analysis: Vasoactive
Liver Congress. J Hepatol 2004; 40:67 medications for the management of acute variceal bleeds.
62. Garcia-Tsao G, Abraldes JG, Berzigotti A, et al: Portal hyper- Aliment Pharmacol Ther 2012; 35:1267–1278
tensive bleeding in cirrhosis: Risk stratification, diagnosis, 79. Gotzsche PC, Hrobjartsson A: Somatostatin analogues for
and management: 2016 practice guidance by the American acute bleeding oesophageal varices. Cochrane Database Syst
Association for the Study of Liver Diseases. Hepatology 2017; Rev 2008; 2008:CD000193
65:310–335 80. D’Amico G, Pagliaro L, Pietrosi G, et al: Emergency sclero-
63. Jung DH, Huh CW, Kim NJ, et al: Optimal endoscopy timing in therapy versus vasoactive drugs for bleeding oesophageal var-
patients with acute variceal bleeding: A systematic review and ices in cirrhotic patients. Cochrane Database Syst Rev 2010;
meta-analysis. Sci Rep 2020; 10:4046 2010:CD002233
64. Greenspoon J, Barkun A: The pharmacological therapy of 81. Garcia-Pagan JC, Caca K, Bureau C, et al; Early TIPS
non-variceal upper gastrointestinal bleeding. Gastroenterol Clin (Transjugular Intrahepatic Portosystemic Shunt) Cooperative
North Am 2010; 39:419–432 Study Group: Early use of TIPS in patients with cirrhosis and
65. Leontiadis GI, Sharma VK, Howden CW: Proton pump inhibi- variceal bleeding. N Engl J Med 2010; 362:2370–2379
tors in acute non-variceal upper gastrointestinal bleeding. J 82. Halabi SA, Sawas T, Sadat B, et al: Early TIPS versus endo-
Gastroenterol Hepatol 2006; 21:1763–1765 scopic therapy for secondary prophylaxis after management
66. Labenz J, Peitz U, Leusing C, et al: Efficacy of primed infusions of acute esophageal variceal bleeding in cirrhotic patients: A
with high dose ranitidine and omeprazole to maintain high meta-analysis of randomized controlled trials. J Gastroenterol
intragastric pH in patients with peptic ulcer bleeding: A pro- Hepatol 2016; 31:1519–1526
spective randomised controlled study. Gut 1997; 40:36–41 83. Savino JA, Cerabona T, Agarwal N, et al: Manipulation of as-
67. Green FW Jr, Kaplan MM, Curtis LE, et al: Effect of acid and citic fluid pressure in cirrhotics to optimize hemodynamic and
pepsin on blood coagulation and platelet aggregation. A pos- renal function. Ann Surg 1988; 208:504–511
sible contributor prolonged gastroduodenal mucosal hemor- 84. Umgelter A, Reindl W, Franzen M, et al: Renal resistive index
rhage. Gastroenterology 1978; 74:38–43 and renal function before and after paracentesis in patients
68. Imhann F, Bonder MJ, Vich Vila A, et al: Proton pump inhibitors with hepatorenal syndrome and tense ascites. Intensive Care
affect the gut microbiome. Gut 2016; 65:740–748 Med 2009; 35:152–156
69. Freedberg DE, Toussaint NC, Chen SP, et al: Proton pump 85. Luca A, Feu F, García-Pagán JC, et al: Favorable effects
inhibitors alter specific taxa in the human gastrointes- of total paracentesis on splanchnic hemodynamics in cir-
tinal microbiome: A crossover trial. Gastroenterology 2015; rhotic patients with tense ascites. Hepatology 1994; 20(1 Pt
149:883–885.e9 1):30–33

Critical Care Medicine www.ccmjournal.org     675

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
 Nanchal et al

86. Mayr U, Karsten E, Lahmer T, et al: Impact of large volume 94. Al-Khafaji A, Elder M, Lebovitz DJ, et al: Protocolized fluid
paracentesis on respiratory parameters including transpul- therapy in brain-dead donors: The multicenter randomized
monary pressure and on transpulmonary thermodilution de- MOnIToR trial. Intensive Care Med 2015; 41:418–426
rived hemodynamics: A prospective study. PLoS One 2018; 95. Bonney GK, Aldersley MA, Asthana S, et al: Donor risk
13:e0193654 index and MELD interactions in predicting long-term graft
87. Mayr U, Fahrenkrog-Petersen L, Batres-Baires G, et al: Large- survival: A single-centre experience. Transplantation 2009;
volume paracentesis effects plasma disappearance rate of 87:1858–1863
Downloaded from http://journals.lww.com/ccmjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCy

indo-cyanine green in critically ill patients with decompensated 96. Gruttadauria S, Vizzini G, Biondo D, et al: Critical use of ex-
liver cirrhosis and intraabdominal hypertension. Ann Intensive tended criteria donor liver grafts in adult-to-adult whole liver
wCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 09/22/2023

Care 2018; 8:78 transplantation: A single-center experience. Liver Transpl

88. De Laet IE, Malbrain M, De Waele JJ: A clinician’s guide to 2008; 14:220–227
management of intra-abdominal hypertension and abdominal 97. Lucidi V, Lemye AC, Baire L, et al: Use of marginal donors for
compartment syndrome in critically ill patients. Crit Care 2020; liver transplantation: A single-center experience within the
24:97 Eurotransplant patient-driven allocation system. Transplant
89. Kirkpatrick AW, Roberts DJ, De Waele J, et al; Pediatric Proc 2007; 39:2668–2671
Guidelines Sub-Committee for the World Society of the 98. Alshamsi F, Alshammari K, Belley-Cote E, et al; GUIDE
Abdominal Compartment Syndrome: Intra-abdominal hyper- Group: Extracorporeal liver support in patients with liver
tension and the abdominal compartment syndrome: Updated failure: A systematic review and meta-analysis of random-
consensus definitions and clinical practice guidelines from ized trials. Intensive Care Med 2020; 46:1–16
the World Society of the Abdominal Compartment Syndrome. 99. Gurusamy KS, Pissanou T, Pikhart H, et al: Methods to
Intensive Care Med 2013; 39:1190–1206 decrease blood loss and transfusion requirements for
90. D’Aragon F, Belley-Cote E, Agarwal A, et al: Effect of corti- liver transplantation. Cochrane Database Syst Rev 2011;
costeroid administration on neurologically deceased organ 2011:CD009052
donors and transplant recipients: A systematic review and 100. Perel P, Roberts I, Ker K: Colloids versus crystalloids for fluid
meta-analysis. BMJ Open 2017; 7:e014436 resuscitation in critically ill patients. Cochrane Database Syst
91. Kotsch K, Ulrich F, Reutzel-Selke A, et al: Methylprednisolone Rev 2013; (2):CD000567
therapy in deceased donors reduces inflammation in the 101. Bampoe S, Odor PM, Dushianthan A, et al: Perioperative
donor liver and improves outcome after liver transplantation: administration of buffered versus non-buffered crystal-
A prospective randomized controlled trial. Ann Surg 2008; loid intravenous fluid to improve outcomes following adult
248:1042–1050 surgical procedures. Cochrane Database Syst Rev 2017;
92. Amatschek S, Wilflingseder J, Pones M, et al: The effect of 9:CD004089
steroid pretreatment of deceased organ donors on liver allo- 102. Semler MW, Self WH, Wanderer JP, et al; SMART Investigators
graft function: A blinded randomized placebo-controlled trial. J and the Pragmatic Critical Care Research Group: Balanced
Hepatol 2012; 56:1305–1309 crystalloids versus saline in critically ill adults. N Engl J Med
93. Kotloff RM, Blosser S, Fulda GJ, et al; Society of Critical Care 2018; 378:829–839
Medicine/American College of Chest Physicians/Association 103. Qureshi SH, Rizvi SI, Patel NN, et al: Meta-analysis of col-
of Organ Procurement Organizations Donor Management loids versus crystalloids in critically ill, trauma and surgical
Task Force: Management of the potential organ donor in the patients. Br J Surg 2016; 103:14–26
ICU: Society of Critical Care Medicine/American College 104. Findlay JY, Jankowski CJ, Vasdev GM, et al: Fast track anes-
of Chest Physicians/Association of Organ Procurement thesia for liver transplantation reduces postoperative ventila-
Organizations Consensus Statement. Crit Care Med 2015; tion time but not intensive care unit stay. Liver Transpl 2002;
43:1291–1325 8:670–675

676     www.ccmjournal.org May 2023 • Volume 51 • Number 5

Copyright © 2023 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.