Analytic Review

Journal of Intensive Care Medicine

1-10
Sepsis and Adrenal Insufficiency © The Author(s) 2023
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/08850666231183396
journals.sagepub.com/home/jic

Cosmo Fowler, MD1, Nina Raoof, MD1,

and Stephen M. Pastores, MD, MACP, FCCP, FCCM1

Abstract
In sepsis, dysregulation of the hypothalamic–pituitary–adrenal axis, alterations in cortisol metabolism, and tissue resistance to
glucocorticoids can all result in relative adrenal insufficiency or critical illness-related corticosteroid insufficiency (CIRCI). The
symptoms and signs of CIRCI during sepsis are nonspecific, generally including decreased mental status, unexplained fever, or
hypotension refractory to fluids, and the requirement of vasopressor therapy to maintain adequate blood pressure. While we
have been aware of this syndrome for over a decade, it remains a poorly understood condition, challenging to diagnose, and asso-
ciated with significantly diverging practices among clinicians, particularly regarding the optimal dosing and duration of corticoste-
roid therapy. The existing literature on corticosteroid use in patients with sepsis and septic shock is vast with dozens of
randomized controlled trials conducted across the past 4 decades. These studies have universally demonstrated reduced duration
of shock, though the effects of corticosteroids on mortality have been inconsistent, and their use has been associated with
adverse effects including hyperglycemia, neuromuscular weakness, and an increased risk of infection. In this article, we aim to
provide a thorough, evidence-based, and practical review of the current recommendations for the diagnosis and management
of patients with sepsis who develop CIRCI, explore the controversies surrounding this topic, and highlight what lies on the hori-
zon as new evidence continues to shape our practice.

Keywords
adrenal insufficiency, sepsis, septic shock, critical illness, corticosteroids, hydrocortisone

Introduction illness-related corticosteroid insufficiency (CIRCI) to more

fully describe the impairment of the HPA axis during critical
Sepsis is defined as life-threatening organ dysfunction caused by a illness.7 CIRCI was defined as inadequate cellular corticoste-
dysregulated host response to infection.1 An intact hypothalamic– roid activity for the severity of the patient’s critical illness.
pituitary–adrenal (HPA) axis is vital in the host response to Although we have been aware of this clinical syndrome for
sepsis.2 In the late 1990s and early 2000s, the concept of relative over a decade, it remains a poorly understood condition, chal-
(as contrasted with absolute) adrenal insufficiency arose to lenging to diagnose, and associated with significant variation
describe patients in septic shock whose HPA reserve was among clinicians regarding the optimal dose and duration of
thought outmatched by the stress of the latter condition, resulting corticosteroid treatment.8
in blunted adrenal corticosteroid production.3 The true incidence In this article, our goal is to provide a thorough, evidence-
of adrenal insufficiency among patients with sepsis and septic based, and practical review of the current recommendations
shock has been difficult to determine. Historically, as with shifting for the diagnosis and management of patients with sepsis who
paradigms, consensus regarding appropriate laboratory testing develop CIRCI, explore the controversies surrounding this
and parameters for its diagnosis has been hard to reach. Annane topic, and highlight what lies on the horizon regarding this quo-
et al3 found that more than half (54%) of 189 patients with tidian clinical entity. Management of critically ill patients on
septic shock had adrenal insufficiency as determined by a short
cosyntropin stimulation test. Similarly, another study reported
that approximately half of the patients with sepsis had concomi-
tant adrenal insufficiency by either baseline cortisol levels or 1
Critical Care Center, Department of Anesthesiology and Critical Care
cosyntropin testing.4 More recently, a study examining the Medicine, Memorial Sloan Kettering Cancer Center, New York, NY, USA
co-occurrence of sepsis and adrenal insufficiency (as determined Received March 31, 2023. Received revised May 12, 2023. Accepted June
by cosyntropin testing) found that in 25% of patients, the 2 con- 5, 2023.
ditions overlapped.5 Another study found a co-occurrence rate of
Corresponding Author:
36.6% (using serum cortisol).6
Stephen M. Pastores, Department of Anesthesiology and Critical Care
In 2008, an international task force convened by the Society Medicine, Memorial Sloan Kettering Cancer Center, 1275 York Avenue
of Critical Care Medicine (SCCM) and the European Society of C-1179, New York, NY, 10065, USA.
Intensive Care Medicine (ESICM) coined the phrase critical Email: pastores@mskcc.org
2 Journal of Intensive Care Medicine 0(0)

chronic steroid therapy is beyond the scope of this review The overall impact of these changes is modulated and amplified
article. by the hepatic GRs through a mechanism that is not fully under-
stood. This complex interplay of abnormal interactions favors a
model of hypercortisolemia in critical illness secondary to
Pathophysiology reduced clearance of cortisol, rather than what was presumed
The Normal HPA Axis to be upregulated production. Furthermore, critical illness is
associated with a shift in GR expression from GRα-dominant
When stressed, humans have developed a response to sustain to one in which GRβ transcription is increased, which has
life.9,10 The HPA axis is a central actor in this response, com- been associated with a blunted response to endogenous gluco-
prised of the hypothalamic paraventricular nucleus (PVN), the corticoids referred to as “tissue resistance.”
anterior pituitary gland, and the adrenal glands. In the hypothala- In both sepsis and acute respiratory distress syndrome
mus, the PVN is responsible for secreting corticotropin-releasing (ARDS), downregulated GRα activity has been hypothesized
hormone (CRH) and arginine vasopressin (AVP), which synerg- and linked, both ex vivo and in vivo, to worsening glucocorticoid
istically incite the release of adrenocorticotropic hormone resistance, morbidity, and mortality. In some experimental
(ACTH) into the circulation.11 ACTH is secreted both in a models, the use of exogenous glucocorticoids has been associated
basal and pulsatile manner. In the physiologic state, this pulsatility with restored GRα expression, function, and resolution of inflam-
is determined primarily by the circadian rhythm with peak levels mation, which may potentially translate into an improved clinical
of cortisol in the morning that decline throughout the day.12,13 In trajectory in the appropriate setting. However, emerging evidence
the adrenal cortex, ACTH induces glucocorticoid synthesis. Most (largely from murine studies) indicates a disparity among different
cortisol remains attached to cortisol-binding globulin (CBG), tissues’ GR profiles and their response to glucocorticoids, which
although an unbound fraction will remain lipophilic, able to trav- clouds the interpretation of a generalized “tissue resistance” and
erse cellular membranes and bind to cytoplasmic glucocorticoid suggests that the above changes of GR expression observed in
receptors (GRs), of which GRα and GRβ are particularly the septic model may be adaptive and protective.2,28
crucial in mediating the stress response.14,15 Free cortisol achieves
varied and powerful immunological and cardiovascular effects
through the binding of GRα. In fact, the cortisol-GRα dimer Clinical Manifestations of CIRCI During
can induce expression or repression of up to 20% of the human Sepsis
genome.16 These effects are directly opposed by GRβ, and ulti-
mately free cortisol inhibits the secretion of both CRH and The symptoms and signs of CIRCI during sepsis are nonspecific
ACTH through negative feedback, framing the HPA axis in a and generally include decreased mental status, unexplained fever,
state of autoregulation.9,10 hypotension refractory to fluids, and the requirement of vasopres-
sor therapy to maintain adequate blood pressure. While hypona-
tremia, hyperkalemia, hypoglycemia, and eosinophilia may be
The HPA Axis During Critical Illness and Sepsis observed, these are generally uncommon, and their presence
may be masked by routine intensive care unit (ICU) care such
During critical illness such as sepsis, the usual organization of
as fluid administration and close electrolyte management.
the HPA axis is upset and replaced by a complex matrix of
axis-independent regulators (Figure 1).17-19 Firstly, it is charac-
terized by an excessive increase in circulating free cortisol Other Confounding States With Abnormal Activation of
which correlates with the severity of the insult.20 Secondly,
the degree of hypercortisolemia is disproportionate to any HPA Axis
increase in cortisol secretion and is completely dissociated Several known and emerging therapies have been associated with
from ACTH levels (ACTH-cortisol dissociation).21,22 Instead, disorders of the HPA axis. Perhaps the most well-known of these
both ACTH and CBG levels fall sharply and see their binding is the sedative etomidate, which inhibits 11β-hydroxylase and,
activity hampered by the hostile environment of critical therefore, the synthesis of cortisol. Though other sedatives can
illness, both mechanisms increasing the fraction of free have this effect, it is only etomidate which is now limited to
cortisol.23 use as a single preintubation induction dose instead of a continu-
Levels of proopiomelanocortin (POMC), the precursor mol- ous infusion for sedation.29 While the one-time use of this agent
ecule of ACTH, are elevated in sepsis through CRH/ has not been definitively linked to increased mortality, retrospec-
AVP-driven POMC expression and impaired processing into tive analyses have shown an excess risk of poor outcomes in the
ACTH, contributing to adrenocortical steroidogenesis and the critically ill. One possible reason for this is the variable effects of
ACTH-cortisol dissociation.24,25 Additionally, the enzymes this drug. The sedative effects of etomidate have an onset of
involved in the breakdown of cortisol in the liver (α-ring reduc- action within 15-20 s with a duration of action of approximately
tases) and kidney (11β-hydroxysteroid dehydrogenase 2) are 2 to 3 min and a favorable hemodynamic profile in comparison to
suppressed in response to inflammation.26 Part of this suppres- other sedatives used in the ICU such as propofol or midazolam.
sion is driven by the accumulation of bile acids, which inhibit However, its effect on the adrenal cortex is much more potent,
the activity of these enzymes through negative feedback.27 with a single dose of etomidate leading to a suppression of cortisol
Fowler et al 3

Figure 1. A newly proposed conceptual framework for adrenocortical function and dysfunction during critical illness.17 Shown here are the
central and peripheral drivers of increased cortisol availability during the acute, subacute, and chronic phases of critical illness. The acute phase,
occurring within minutes to hours after the initial insult, is mostly characterized by a central, ACTH-driven rise in cortisol and a peripherally
driven (by decreases in cortisol binding to cortisol binding proteins, in addition to GR-driven effects) further rise in free cortisol. By contrast,
the subacute phase (occurring hours to days after the initial insult) is marked by a central suppression driven by feedback inhibition with
sustained peripherally driven elevated total cortisol and free cortisol. During the prolonged phase, extending beyond several weeks of critical
illness, the sustained central suppression of the HPA axis caused by feedback inhibition driven by sustained elevation of free cortisol and by
elevated bile acids and/or drugs could lead to central hypoadrenalism.
Abbreviations: ACTH, adrenocorticotropic hormone; GR, glucocorticoid receptor; CBG, cortisol-binding globulin; CRH, corticotropin-releasing hormone; HPA,
hypothalamic–pituitary–adrenal.

synthesis for up to 72 h. Additionally, it can interfere with other corticosteroid therapy in contrast to the rapidly tapering regimens
steroid-mediated processes.2 The mixed results of several meta- often used for CIRCI.32 Another novel cancer treatment associ-
analyses investigating this agent have left the debate over its ated with a sepsis-like presentation is chimeric antigen receptor
use in the critically ill ultimately unresolved.30 (CAR)-T cell therapy. CAR-T can induce a hyperinflammatory
Immune checkpoint inhibitor (ICI) regimens and cellular ther- state known as cytokine release syndrome (CRS), which can be
apies, which modulate T-cell function to fight cancer cells, can difficult to distinguish from sepsis, with fever and hypotension
also ignite immune-related adverse events including hypophysitis being hallmarks of the condition. Current management of CRS
(inflammation of the pituitary gland), which can lead to a central relies on suppression of the powerful inflammatory response
(or “secondary”) inhibition of ACTH release, and primary adrenal this therapy engenders, usually with specific cytokine antagonists
insufficiency, in which the synthesis of glucocorticoids by the (eg, tocilizumab and interleukin-6 receptor antagonist) and often
adrenal glands is inhibited. Given the increased utilization of alongside high doses of glucocorticoids.33,34
ICIs worldwide, it is important for intensivists to be aware of
these possible mimics or confounding pathologies of CIRCI. A
recent study showed that particular ICI combinations may result Laboratory Testing
in relatively high rates of adrenal insufficiency (almost one-third Patients with CIRCI typically have normal or low plasma total
of patients with nivolumab/ipilimumab combination therapy).31 cortisol levels and normal to low plasma ACTH. However,
The use of ICIs may also result in a syndrome requiring long-term there is no single test that can reliably diagnose CIRCI in patients
4 Journal of Intensive Care Medicine 0(0)

with sepsis. The 2008 CIRCI guidelines suggested that the diag- For most of these landmark studies, IV hydrocortisone was
nosis of CIRCI is best made by an increase in total serum cortisol used (with or without fludrocortisone) at a dose of 200 mg
of <9 µg/dL after i.v. cosyntropin (synthetic ACTH) 250 µg daily in divided doses (50 mg IV every 6 h) or continuous infu-
administration or a random total cortisol of <10 µg/dL.7 ACTH sion for at least 7 days without a taper schedule. It is notable that
stimulation testing likely has limited applicability in critically ill only the RCTs that used a combination of hydrocortisone and
patients, and the 2017 task force could not reach a consensus fludrocortisone showed a mortality benefit.44,46
on whether the test was superior to random cortisol.19 It is impor- Though each of these RCTs differed in various ways, the
tant to note that ACTH stimulation testing only tests adrenal primary outcome under consideration was 28- or 90-day mor-
responsiveness to ACTH and is not a test of the HPA axis.35 tality, and a secondary outcome that they all investigated was
The 2017 task force was not in favor of using plasma-free cortisol time to shock reversal. Some studies also evaluated days on
or salivary gland cortisol level over plasma total cortisol level as mechanical ventilation (MV), the occurrence of superinfection,
the evidence for this was of poor quality (low-quality observa- recurrent shock, gastrointestinal (GI) bleeding, neuromuscular
tional studies with inconsistent findings). Similarly, assessing weakness, and ICU and hospital length of stay. While these
the hemodynamic response to hydrocortisone to diagnose landmark studies did not concur on the benefit of corticosteroids
CIRCI was not recommended. Although the low-dose (1 µg) regarding mortality benefit, they all found a shorter time to
ACTH stimulation test may have similar diagnostic accuracy shock reversal with corticosteroid therapy. In fact, the largest
and accuracy in predicting vasopressor dependency and clinical study (ADRENAL)47 also noted a shorter time to ICU dis-
outcomes in septic patients, it requires some preparation and charge. Considering the volume of patients admitted to the
offers no advantage over the high-dose ACTH test.36 ICU with sepsis, if this carries over into day-to-day ICU care,
it might have a beneficial impact on ICU staffing and workflow
in addition to patient-level outcomes.
One way to make sense of these studies at the bedside is to
Treatment With Corticosteroids:
compare and contrast them. The 2 studies conducted in France
Who, When, and How to Treat? (Annane’s original study in 2002 and APROCCHSS in 2018)
Over the past 4 decades, numerous randomized controlled trials both used a combination of hydrocortisone with fludrocortisone
(RCTs) have been conducted evaluating the role of corticoste- in contrast to CORTICUS45 and ADRENAL. Additionally,
roids in a wide variety of patients with sepsis and septic only the 2 French studies demonstrated a survival benefit
shock. These include adult, pregnant, and pediatric patients with the coadministration of hydrocortisone plus fludrocorti-
with and without ARDS who were treated with varying sone. This coincidence has led to the hypothesis of fludrocorti-
doses, dosing strategies, formulations, and durations of cortico- sone coadministration being instrumental in these 2 studies’
steroid therapy. In the 1980s, RCTs using high-dose corticoste- intervention arms securing a survival advantage, though this
roid regimens (>400 mg/day hydrocortisone or equivalent for seems unlikely, as the mineralocorticoid effect that might be
<3 days) in patients with septic shock demonstrated no benefi- anticipated from 50 μg of fludrocortisone is already readily
cial treatment effect and even suggested that treatment may achieved by 200 mg of hydrocortisone (regardless of the
increase mortality.37 Use of lower dose corticosteroids (referred enteral bioavailability of fludrocortisone in the vasopressor-
to as “stress dose” steroids) was revived in the late 1990s when dependent, critically ill patient).17 Another hypothesis to
2 small-sized RCTs showed faster reversal of shock and explain the outcome discrepancy between the 2 Annane
improved mortality.38,39 studies and the CORTICUS/ADRENAL trials notes the
While different glucocorticoid agents have been trialed in French trials had higher acuity patients (mortality among
sepsis and septic shock, a recent systematic review and meta- patients randomized in APROCCHSS was almost double that
analysis of glucocorticoid therapy in septic shock compared of ADRENAL). Importantly, all 4 studies showed a shorter
18 studies using hydrocortisone, 2 that used methylpredniso- time to reversal of shock. However, this alone (discontinuation
lone (n = 158 patients) and 1 with dexamethasone (n = 29 of vasopressor therapy) cannot be seen as a meaningful patient-
patients).40-43 As the greater part of the evidence examining centered outcome. We can contrast this with the decreased use
corticosteroid use in sepsis is derived from trials that utilized of blood transfusions and shorter time on the ventilator identi-
hydrocortisone, our recommendations and those of the profes- fied by the ADRENAL trial.
sional societies involve the use of this agent. The use of corticosteroids in sepsis and septic shock is not
Table 1 lists the 4 landmark RCTs addressing the efficacy without risk. The CORTICUS trial showed that patients
and safety of corticosteroids in patients with sepsis and septic treated with corticosteroids had an increased risk of superinfec-
shock over the past 20 years. Studies that included steroids as tion and recurrent septic shock. Both of these findings may have
part of combination therapy with nonsteroidal agents significant downstream effects on patients. Finally, all of the
(eg, vitamin C with hydrocortisone) were not included. In addi- landmark trials found that corticosteroid therapy increased the
tion to appreciating the nuances of each RCT, it is worth appre- risk of hyperglycemia, a known risk factor for increased mor-
ciating that this body of literature represents some of the most bidity and mortality in the critically ill.48 How, then, can we
robust data gathered in the critical care population across 3 interpret this data in the face of uncertain and at times conflict-
continents. ing findings? It seems most appropriate to consider these studies
 Table 1. Four Landmark RCTs Addressing the Efficacy and Safety of Corticosteroids in Patients With Sepsis and Septic Shock Over the Past 20 Years.

Secondary
Study (year) Country(s) Population Intervention Controls Primary outcome(s) outcome(s) Results Adverse effects

Annane et al France 299 adult septic Hydrocortisone 50 Placebo 28-day mortality in ICU and hospital Mortality (28-day, Similar risk of
(2002)44 shock patients on mg i.v. q.6 h + patients with mortality, 1-year ICU, and superinfection, GI
MV with fludrocortisone 50 relative adrenal mortality, between hospital) lower bleeding, and
documented/ μg p.o. q.24 h for 7 insufficiency patients with/ and shorter psychiatric
suspected days (cortisol level without response time to shock disturbance
infection and increase of ≤9 μg/ to corticotropin reversal in the between arms
signs of shock dL in tests and time to intervention
corticotropin shock reversal arm
test)
CORTICUS Austria, Belgium, 499 adult patients, Hydrocortisone 50 Placebo 28-day mortality ICU and hospital No mortality Increased risk of
Sprung et al France, with evidence of mg i.v. q 6 h for 5 mortality, 1-year difference, a superinfection,
(2008)45 Germany, infection, 2/4 days, then tapered mortality, time to shorter time to hyperglycemia,
Israel, Italy, SIRS criteria, and over 6 days shock reversal shock reversal hypernatremia, and
Netherlands, shock within the in the weakness in the
Portugal, and previous 72 h intervention intervention arm
the United primarily with arm
Kingdom septic shock
APROCCHSS France 1241 adult patients Hydrocortisone 50 Placebo 90-day mortality Time to shock Mortality is lower Increased risk of
Annane with proven/ mg i.v. q 6 h + reversal, time to and shorter hyperglycemia in
et al probable septic fludrocortisone 50 weaning from MV time to shock the intervention
(2018)46 shock ≤ 24 h μg p.o. q. 24 h for 7 reversal in the arm, similar risk of
prior to days intervention GI bleeding, and
enrollment arm superinfection
ADRENAL Australia, New 3658 adult septic Hydrocortisone 200 Placebo 90-day mortality 28-day mortality, No mortality Increased risk of
Venkatesh Zealand, United shock patients on mg i.v. q 24 h time to shock difference, hyperglycemia,
et al Kingdom, MV with continuously reversal/ shorter time to hypernatremia,
(2018)47 Denmark, and documented/ infused for 7 days recurrence, time shock reversal, encephalopathy,
Saudi Arabia suspected (or to ICU to wean from MV, and ICU and myopathy,
infection, 2/4 discharge if hospital and ICU discharge in the similar risk of new
SIRS criteria, shorter) LOS, new intervention bacteremia
≥4 h of infection arm
vasopressors

Abbreviations: RCT, randomized controlled trial; MV, mechanical ventilation; ICU, intensive care unit; GI, gastrointestinal; MI, myocardial infarction; LOS, length of stay.

5
6 Journal of Intensive Care Medicine 0(0)

in the context of the existing meta-analyses and national society Weakness

practice guidelines.
Sepsis and other forms of critical illness requiring ICU admis-
A systematic review with meta-analysis and trial sequential
sion are associated with profound deconditioning in survivors.
analysis of 22 RCTs including 7297 participants concluded that
This is secondary to the marked catabolism of the critically ill
in adult patients with septic shock treated with corticosteroids in
state, with an approximate loss of muscle mass of 2% to 4%
a dose of <500 mg per day of hydrocortisone (or equivalent)
per day of ICU stay,59 but also the toll of therapies used to
compared to placebo or usual care, there was a reduced duration
combat this state. Glucocorticoids potentiate protein degrada-
of shock, MV, and ICU stay, but no significant change in short-
tion and loss of lean muscle mass,60 and prolonged use may
and longer-term mortality, and an increased rate of adverse
increase the risk and severity of post-ICU syndrome with its
events.40 Similarly, an updated systematic review and meta-
attendant negative effects on quality of life.61 However, a
analysis of 42 RCTs including 10 194 patients concluded that
review article found that while many case reports and series
corticosteroids possibly result in a small reduction in mortality
in the 1970s and 1980s reported weakness in patients after
(2% absolute risk reduction) and hospital and ICU length of
receiving high-dose corticosteroids, this finding was not repli-
stay, but at the expense of increased rates of hyperglycemia,
cated in prospective studies, nor in contemporary randomized
hypernatremia, and neuromuscular weakness.49 Certainly, the
trials that utilized lower doses of corticosteroids and improved
increased risk of neuromuscular weakness may have an
glycemic control.62 Corticosteroids were not associated with
outsize effect on the patient’s quality of life.
increased duration of mechanical ventilatory support, one of
Given the volume of unclear and conflicting evidence, it is
the most significant side effects of weakness in the critically
vital to know what our professional organizations recommend
ill.62
to clinicians at the bedside. The 2017 SCCM/ESICM guidelines
suggested the use of <400 mg/day of IV hydrocortisone for ≥3
days at full dose in adult patients with septic shock that is not Delirium
responsive to fluid and moderate- to high-dose vasopressor
therapy.19 Most recently, the 2021 Surviving Sepsis Guidelines Glucocorticoid use may be associated with hyperarousal and
suggested using IV corticosteroids (IV hydrocortisone 200 mg/ sleep disturbance, behavioral changes, delirium, and frank psy-
day given as 50 mg IV every 6 h or as a continuous infusion) chosis.63 As with hyperglycemia, ICU delirium is a well-
for adults with septic shock and an ongoing requirement for vaso- established independent risk factor for morbidity and mortality
pressor therapy (≥0.25 mcg/kg/min of norepinephrine or epi- in critically ill patients.64 However, among the landmark RCTs,
nephrine for at least 4 h), based on moderate quality evidence.50 only the ADRENAL trial showed an increased rate of enceph-
Coadministration of fludrocortisone was not addressed. On the alopathy in septic patients who received IV hydrocortisone.47
basis of the evidence reviewed, professional consensus, and our More broadly, the biological underpinnings of delirium have
own practice, we recommend the administration of 200 mg/day yet to be fully understood, much less the role of endogenous
of hydrocortisone IV in divided doses (or by infusion) to patients cortisol in this process. Exogenous corticosteroids have been
with ongoing vasopressor requirement outlined in the Surviving associated with both decreased and increased risk of delirium.
Sepsis Guidelines, for at least 3 to 5 days with or without taper. The Hydrocortisone for Prevention of Septic Shock
A recent retrospective study showed that abrupt discontinuation (HYPRESS) study unexpectedly found lower rates of delirium
of corticosteroids was associated with a lower likelihood of vaso- in patients treated with hydrocortisone than in those who
pressor reinitiation when compared with taper discontinuation.51 received a placebo.65 Thus, the current literature does not
In the event of shock recrudescence, the decision to reinstitute support a clearly positive or negative association between glu-
hydrocortisone should be tailored to the specific clinical circum- cocorticoid use and delirium.64
stances. Our recommendations regarding corticosteroid adminis-
tration in various systemic inflammatory processes are
Infection
summarized in Table 2.
Concern for promoting infection risk has long been a theoretical
detractor to the use of corticosteroids in patients with sepsis and
Adverse Effects of Corticosteroids septic shock. However, only the CORTICUS study found
increased rates of superinfection among the septic patients
Hyperglycemia who received IV hydrocortisone.45
Most landmark trials examining the use of adjunctive glucocor-
ticoids in sepsis and septic shock showed an excess of hypergly-
cemia events in the corticosteroid arm. This occurs through
GI Bleeding
several mechanisms, such as glucocorticoids’ promotion of glu- As with the risk of infection, the inhibition of gastroprotective
coneogenesis, inhibition of insulin secretion and glucose prostaglandins and the risk of GI hemorrhage in already-
uptake, and increased protein degradation.58 If this dysregula- predisposed critically ill patients has been of concern with glu-
tion is not appropriately managed, hyperglycemia has been cocorticoid use in sepsis for decades. In aggregate, hospitalized
clearly associated with increased ICU mortality. patients receiving glucocorticoids have been found to have a
Fowler et al 7

Table 2. Administration of Corticosteroids to Patients With Various Systemic Inflammatory Processes and Sepsis (Who, When, and How?).

Patient Dose Duration Notes

Inpatients with SIRS Do not treat

Inpatients with sepsis
Inpatients with meningitis Dexamethasone 0.15 mg/kg q.6h 4 days Early, adjuvant glucocorticoids in bacterial
beginning shortly before or at the meningitis improve morbidity and mortality
same time as the initiation of (primarily Salmonella pneumoniae), particularly
antibiotics 52 in developed regions, as opposed to those
developing53,54
Inpatients with CAP Treat for severe CAP; however, there There is no The 2019 ATS and IDSA guidelines recommend
are no consensus statements consensus against adjunctive glucocorticoids55
regarding the dose statement A 2021 systematic review and meta-analysis
regarding the showed a lower incidence in need for MV56
duration A 2023 RCT showed decreased 28-day
mortality in ICU patients with severe CAP
who received IV hydrocortisone 200 mg daily
for either 4 or 8 days depending on clinical
improvement followed by tapering for a total
of 8 or 14 days compared to placebo57
Inpatients with early Dexamethasone 20 mg IV daily tapered 5 days up to 10 days In unresolving ARDS (7-21 days)18
non-COVID-19 ARDS to 10 mg IV daily unless extubated Methylprednisolone 2 mg/kg IV bolus, then
requiring invasive MV • Day 1-14: 2 mg/kg/day divided q. 6h
Methylprednisolone 1 mg/kg IV bolus Up to 3 days or 28 • Day 15-21: 1 mg/kg/day
followed by 1 mg/kg/day followed by days • Day 22-28: 0.5 mg/kg/day
slow taper • Day 29-30: 0.25 mg/kg/day
• Day 31-32: 0.125 mg/kg/day

If extubated prior to day 14, then advanced to

day 15 of regimen drug therapy

Critically ill patients with Hydrocortisone 200 mg IV daily as a After >4 h of requiring norepinephrine or
septic shock continuous infusion or 50 mg q.6h epinephrine ≥0.25 mcg/kg/min50

Abbreviations: CAP, community-acquired pneumonia; ARDS, acute respiratory distress syndrome; MV, mechanical ventilation; ICU, intensive care unit.

higher risk of GI bleed and perforation, but the same meta- response to sepsis. Among these is the endotype A/B schema,
analysis found the subset receiving corticosteroids for sepsis which refers to differing patterns of genes expressed.68 Adults
or septic shock not to have an increased risk of GI bleeding.66 with endotype A1 had a 5-fold increased mortality between
More importantly, at the commonly prescribed dose of gluco- the ages of 18 and 40 years, although these differences dimin-
corticoids for septic shock, no major trials have established a ished by age 60. Other ways to characterize these patients have
direct link between steroid use and increased risk of GI been investigated, including stratification by proteomics and
bleeding. transcriptomics.69,70 One small study has suggested that there
may be a group of critically ill patients with sepsis who are
harmed by the administration of steroids. A post hoc analysis
Future Research Directions of the Vasopressin versus Norepinephrine as Initial Therapy
As regards ongoing investigation into how best to manage crit- in Septic Shock (VANISH) RCT,71 which enrolled patients in
ically ill patients with sepsis, a phase II, multicenter RCT67 is shock and randomized them to receive norepinephrine or vaso-
currently underway to address one of the least understood pressin followed by hydrocortisone or placebo was enlighten-
aspects of corticosteroid therapy—the use of mineralocorti- ing.72 It identified 2 transcriptomic sepsis response signatures
coids. In this study, the 200 mg IV daily dose of hydrocortisone (SRSs), one relatively immune-suppressed (SRS1) and the
will be compared to 3 different doses of fludrocortisone to iden- other, immuno-competent (SRS2) with different outcomes
tify the possible impact of this combination therapy on a variety after corticosteroid therapy. Patients with the immunocompe-
of primary and secondary outcomes including mortality, length tent SRS2 endotype who were treated with corticosteroids
of stay, shock recurrence, and ventilator-free days. had higher mortality than those who were given a placebo but
A way to reframe the adrenal insufficiency in the sepsis nar- not those with the SRS1 endotype. While these findings are
rative is to consider whether there may exist different pheno- promising, further validation studies are needed, and it is too
types of adrenal response to sepsis. Various strategies of early to consider them useful when managing patients at the
genome analysis have been proposed to categorize the human bedside.
8 Journal of Intensive Care Medicine 0(0)

Another novel multicenter study currently recruiting patients ORCID iD

in France has the goal of enrolling patients with sepsis and Stephen M. Pastores https://orcid.org/0000-0002-8850-7062
septic shock, identifying the particular biomarker profile or phe-
notype they express, and randomizing them to receive either References
hydrocortisone and fludrocortisone or placebo. The goal of
1. Singer M, Deutschman CS, Seymour CW, et al. The third interna-
this study will be to identify any subsets of patients who may
tional consensus definitions for sepsis and septic shock (sepsis-3).
benefit or be harmed by treatment with corticosteroids. The JAMA. 2016;315(8):801-810. doi:10.1001/jama.2016.0287
primary outcome is mortality at 90 days or persistent organ dys- 2. Van den Berghe G, Téblick A, Langouche L, Gunst J. The hypothal-
function, and secondary outcomes will include vasopressor-free amus–pituitary–adrenal axis in sepsis- and hyperinflammation-
days. Enrollment is expected to be completed by 2025.73 induced critical illness: gaps in current knowledge and future trans-
Finally, as with any other study involving the critically ill, it lational research directions. EBioMedicine. 2022;84:104284. doi:10.
is important that we consider our research goals. The most 1016/j.ebiom.2022.104284
investigated outcomes of critical care trials are mortality, 3. Annane D, Sébille V, Troché G, Raphaël J-C, Gajdos P, Bellissant
infection-related outcomes, and ventilation-related outcomes.74 E. A 3-level prognostic classification in septic shock based on cor-
It is worth considering whether alternative patient-centered out- tisol levels and cortisol response to corticotropin. JAMA.
comes such as discharge home or to a skilled nursing facility or 2000;283(8):1038-1045. doi:10.1001/jama.283.8.1038
4. Ho HC, Chapital AD, Yu M. Hypothyroidism and adrenal insuf-
measures of functional status may be more meaningful to most
ficiency in sepsis and hemorrhagic shock. Arch Surg.
clinicians and their patients. 2004;139(11):1199-1203. doi:10.1001/archsurg.139.11.1199
5. Hashemi-Madani N, Miri M, Emami Z, Barati M, Golgiri F.
Adrenal insufficiency in septic patients admitted to intensive
Conclusions care unit: prevalence and associated factors. Med J Islam Repub
Relative adrenal insufficiency or CIRCI occurs frequently in Iran. 2021;35:154.
patients with sepsis due to dysregulation of the HPA axis, alter- 6. Swain K, Nanda S, Jayasingh SC, Routray SS, Swain A, Sahoo P.
Occurrence of adrenal suppression in patients having sepsis in
ations in cortisol metabolism, and tissue resistance to glucocor-
Indian population and impact of corticosteroid supplementation
ticoids. Several large meta-analyses of RCTs have demonstrated
on its overall survival. Int J Recent Surg Med Sci. 2021;7(2):80-
that administration of corticosteroids results in reduced duration 84. doi:10.1055/s-0041-1731916
of septic shock, although the effects on mortality have been 7. Marik PE, Pastores SM, Annane D, et al. Recommendations for
inconsistent, and their use is associated with potential adverse the diagnosis and management of corticosteroid insufficiency in
effects, most notably hyperglycemia. The optimal dose, timing critically ill adult patients: consensus statements from an interna-
of initiation, and duration of corticosteroid therapy for septic tional task force by the American College of Critical Care
shock remain uncertain. The majority of critical care clinicians Medicine. Crit Care Med. 2008;36(6):1937-1949. doi:10.1097/
typically administer 200 mg/day of hydrocortisone IV in CCM.0b013e31817603ba
divided doses (or by infusion) for at least 3 to 5 days or until 8. Chan ED, Chan MM, Chan MM, Marik PE. Use of glucocorti-
the vasopressor dose starts decreasing or is weaned off. Some cli- coids in the critical care setting: science and clinical evidence.
nicians may consider adding fludrocortisone at 0.05 mg (50 µg) Pharmacol Ther. 2020;206:107428. doi:10.1016/j.pharmthera.
2019.107428
orally once daily. In the near future, we can look forward to
9. Cain DW, Cidlowski JA. Immune regulation by glucocorticoids.
updated guidelines from the SCCM for more comprehensive Nat Rev Immunol. 2017;17(4):233-247. doi:10.1038/nri.2017.1
guidance on dosing and duration of corticosteroid therapy in 10. Russell G, Lightman S. The human stress response. Nat Rev
patients with septic shock. Endocrinol. 2019;15(9):525-534. doi:10.1038/s41574-019-0228-0
11. Timmermans S, Souffriau J, Libert C. A general introduction to
Authors Contribution glucocorticoid biology. Front Immunol. 2019;10:1545. doi:10.
3389/fimmu.2019.01545
All authors contributed to the conception and design, acquisition, anal-
12. Walker JJ, Spiga F, Gupta R, Zhao Z, Lightman SL, Terry JR.
ysis, and interpretation of data for the work and final approval of the
Rapid intra-adrenal feedback regulation of glucocorticoid synthe-
revised manuscript.
sis. J R Soc Interface. 2015;12(102):20140875. doi:10.1098/rsif.
2014.0875
Declaration of Conflicting Interests 13. Boonen E, Meersseman P, Vervenne H, et al. Reduced nocturnal
The author(s) declared no potential conflicts of interest with respect to ACTH-driven cortisol secretion during critical illness. Am J
the research, authorship, and/or publication of this article. Physiol Endocrinol Metab. 2014;306(8):E883-E892. doi:10.
1152/ajpendo.00009.2014
14. Nixon M, Upreti R, Andrew R. 5α-reduced glucocorticoids: a
Funding story of natural selection. J Endocrinol. 2012;212(2):111-127.
The author(s) disclosed receipt of the following financial support for doi:10.1530/JOE-11-0318
the research, authorship, and/or publication of this article: This work 15. Wepler M, Preuss JM, Merz T, et al. Impact of downstream effects
was supported, in part, by the Craig Thompson, Core Grant (P30 of glucocorticoid receptor dysfunction on organ function in criti-
CA008748) and the Department of Anesthesiology and Critical Care cal illness-associated systemic inflammation. Intensive Care
Medicine, Memorial Sloan Kettering Cancer Center, New York, NY. Med Exp. 2020;8(Suppl 1):37. doi:10.1186/s40635-020-00325-z
Fowler et al 9

16. Galon J, Franchimont D, Hiroi N, et al. Gene profiling reveals analysis. JAMA Oncol. 2018;4(2):173-182. doi:10.1001/
unknown enhancing and suppressive actions of glucocorticoids jamaoncol.2017.3064
on immune cells. FASEB J. 2002;16(1):61-71. doi:10.1096/fj. 32. Barroso-Sousa R, Ott PA, Hodi FS, Kaiser UB, Tolaney SM, Min
01-0245com L. Endocrine dysfunction induced by immune checkpoint inhibi-
17. Téblick A, Peeters B, Langouche L, Van den Berghe G. Adrenal tors: practical recommendations for diagnosis and clinical man-
function and dysfunction in critically ill patients. Nat Rev agement. Cancer. 2018;124(6):1111-1121. doi:10.1002/cncr.
Endocrinol. 2019;15(7):417-427. doi:10.1038/s41574-019-0185-7 31200
18. Annane D, Pastores SM, Arlt W, et al. Critical illness-related cor- 33. Morris EC, Neelapu SS, Giavridis T, Sadelain M. Cytokine
ticosteroid insufficiency (CIRCI): a narrative review from a multi- release syndrome and associated neurotoxicity in cancer immuno-
specialty task force of the Society of Critical Care Medicine therapy. Nat Rev Immunol. 2022;22(2):85-96. doi:10.1038/
(SCCM) and the European Society of Intensive Care Medicine s41577-021-00547-6
(ESICM). Intensive Care Med. 2017;43(12):1781-1792. doi:10. 34. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus
1007/s00134-017-4914-x grading for cytokine release syndrome and neurologic toxicity
19. Annane D, Pastores SM, Rochwerg B, et al. Guidelines for the associated with immune effector cells. Biol Blood Marrow
diagnosis and management of critical illness-related corticosteroid Transplant. 2019;25(4):625-638. doi:10.1016/j.bbmt.2018.12.
insufficiency (CIRCI) in critically ill patients (Part I): Society of 758
Critical Care Medicine (SCCM) and European Society of 35. Bornstein SR, Allolio B, Arlt W, et al. Diagnosis and treatment of
Intensive Care Medicine (ESICM) 2017. Intensive Care Med. primary adrenal insufficiency: an endocrine society clinical prac-
2017;43(12):1751-1763. doi:10.1007/s00134-017-4919-5 tice guideline. J Clin Endocrinol Metab. 2016;101(2):364-389.
20. Widmer IE, Puder JJ, König C, et al. Cortisol response in relation doi:10.1210/jc.2015-1710
to the severity of stress and illness. J Clin Endocrinol Metab. 36. Alvarez-Payares JC, Bello-Simanca JD, De La Peña-Arrieta EDJ,
2005;90(8):4579-4586. doi:10.1210/jc.2005-0354 et al. Common pitfalls in the interpretation of endocrine tests.
21. Boonen E, Vervenne H, Meersseman P, et al. Reduced cortisol Front Endocrinol (Lausanne). 2021;12:727628. doi:10.3389/
metabolism during critical illness. N Engl J Med. fendo.2021.727628
2013;368(16):1477-1488. doi:10.1056/NEJMoa1214969 37. Cronin L, Cook DJ, Carlet J, et al. Corticosteroid treatment for
22. Raff H, Biru N, Reisinger N, Kramer DJ. Dissociation of ACTH sepsis: a critical appraisal and meta-analysis of the literature.
and cortisol in septic and non-septic ICU patients. Endocrine. Crit Care Med. 1995;23(8):1430-1439. doi:10.1097/
2017;55(1):307-310. doi:10.1007/s12020-016-1034-2 00003246-199508000-00019
23. Peeters B, Meersseman P, Vander Perre S, et al. Adrenocortical 38. Bollaert P-E, Charpentier C, Levy B, Debouverie M, Audibert G,
function during prolonged critical illness and beyond: a prospective Larcan A. Reversal of late septic shock with supraphysiologic
observational study. Intensive Care Med. 2018;44(10):1720-1729. doses of hydrocortisone. Crit Care Med. 1998;26(4):645-650.
doi:10.1007/s00134-018-5366-7 39. Briegel J, Forst H, Haller M, et al. Stress doses of hydrocorti-
24. Téblick A, Vander Perre S, Pauwels L, et al. The role of sone reverse hyperdynamic septic shock: a prospective, ran-
pro-opiomelanocortin in the ACTH-cortisol dissociation of sepsis. domized, double-blind, single-center study. Crit Care Med.
Crit Care. 2021;25(1):65. doi:10.1186/s13054-021-03475-y 1999;27(4):723-732.
25. Téblick A, Gunst J, Van den Berghe G. Critical illness-induced 40. Rygård SL, Butler E, Granholm A, et al. Low-dose corticosteroids
corticosteroid insufficiency: what it is not and what it could be. for adult patients with septic shock: a systematic review with
J Clin Endocrinol Metab. 2022;107(7):2057-2064. doi:10.1210/ meta-analysis and trial sequential analysis. Intensive Care Med.
clinem/dgac201 2018;44(7):1003-1016. doi:10.1007/s00134-018-5197-6
26. Van den Berghe G. Adrenal function/dysfunction in critically ill 41. Torres A, Sibila O, Ferrer M, et al. Effect of corticosteroids on
patients: a concise narrative review of recent novel insights. treatment failure among hospitalized patients with severe
J Anesth. 2021;35(6):903-910. doi:10.1007/s00540-021-02977-x community-acquired pneumonia and high inflammatory response:
27. Jenniskens M, Langouche L, Vanwijngaerden YM, Mesotten D, a randomized clinical trial. JAMA. 2015;313(7):677-686. doi:10.
Van den Berghe G. Cholestatic liver (dys)function during sepsis 1001/jama.2015.88
and other critical illnesses. Intensive Care Med. 2016;42(1):16- 42. Deng QM, Shang D, Wan XY. [Effects of low-dose glucocorti-
27. doi:10.1007/s00134-015-4054-0 coid on renal function and prognosis in patients with sepsis].
28. Téblick A, Van Dyck L, Van Aerde N, et al. Impact of duration of Zhongguo Wei Zhong Bing Ji Jiu Yi Xue. 2011;23(3):183-184.
critical illness and level of systemic glucocorticoid availability on 43. Cicarelli DD, Vieira JE, Benseñor FE. Early dexamethasone treat-
tissue-specific glucocorticoid receptor expression and actions: a ment for septic shock patients: a prospective randomized clinical
prospective, observational, cross-sectional human and two transla- trial. Sao Paulo Med J. 2007;125(4):237-241. doi:10.1590/
tional mouse studies. EBioMedicine. 2022;80:104057. doi:10. s1516-31802007000400009
1016/j.ebiom.2022.104057 44. Annane D, Sébille V, Charpentier C, et al. Effect of treatment with
29. Besnier E, Clavier T, Compere V. The hypothalamic–pituitary– low doses of hydrocortisone and fludrocortisone on mortality in
adrenal axis and anesthetics: a review. Anesth Analg. patients with septic shock. JAMA. 2002;288(7):862-871. doi:10.
2017;124(4):1181-1189. doi:10.1213/ANE.0000000000001580 1001/jama.288.7.862
30. Albert SG, Sitaula S. Etomidate, adrenal insufficiency and mortality 45. Sprung CL, Annane D, Keh D, et al. Hydrocortisone therapy for
associated with severity of illness: a meta-analysis. J Intensive Care patients with septic shock. N Engl J Med. 2008;358(2):111-124.
Med. 2021;36(10):1124-1129. doi:10.1177/0885066620957596 doi:10.1056/NEJMoa071366
31. Barroso-Sousa R, Barry WT, Garrido-Castro AC, et al. Incidence 46. Annane D, Renault A, Brun-Buisson C, et al. Hydrocortisone plus
of endocrine dysfunction following the use of different immune fludrocortisone for adults with septic shock. N Engl J Med.
checkpoint inhibitor regimens: a systematic review and meta- 2018;378(9):809-818. doi:10.1056/NEJMoa1705716
10 Journal of Intensive Care Medicine 0(0)

47. Venkatesh B, Finfer S, Cohen J, et al. Adjunctive glucocorticoid stakeholders’ conference. Crit Care Med. 2012;40(2):502-509.
therapy in patients with septic shock. N Engl J Med. doi:10.1097/CCM.0b013e318232da75
2018;378(9):797-808. doi:10.1056/NEJMoa1705835 62. Wilcox SR. Corticosteroids and neuromuscular blockers in devel-
48. Corathers SD, Falciglia M. The role of hyperglycemia in acute opment of critical illness neuromuscular abnormalities: a historical
illness: supporting evidence and its limitations. Nutrition. review. J Crit Care. 2017;37:149-155. https://doi.org/10.1016/j.
2011;27(3):276-281. https://doi.org/10.1016/j.nut.2010.07.013 jcrc.2016.09.018
49. Rochwerg B, Oczkowski SJ, Siemieniuk RAC, et al. 63. Cole JL. Steroid-Induced sleep disturbance and delirium: a
Corticosteroids in sepsis: an updated systematic review and meta- focused review for critically ill patients. Fed Pract.
analysis. Crit Care Med. 2018;46(9):1411-1420. doi:10.1097/ 2020;37(6):260-267.
ccm.0000000000003262 64. Hill AR, Spencer-Segal JL. Glucocorticoids and the brain after
50. Evans L, Rhodes A, Alhazzani W, et al. Surviving sepsis cam- critical illness. Endocrinology. 2021;162(3):bqaa242. doi:10.
paign: international guidelines for management of sepsis and 1210/endocr/bqaa242
septic shock 2021. Intensive Care Med. 2021;47(11):1181-1247. 65. Keh D, Trips E, Marx G, et al. Effect of hydrocortisone on devel-
doi:10.1007/s00134-021-06506-y opment of shock among patients with severe sepsis: the
51. Carabetta S, Allen B, Cannon C, Johnson T. Abrupt discontin- HYPRESS randomized clinical trial. JAMA. 2016;316(17):1775-
uation versus taper of hydrocortisone in patients with septic 1785. doi:10.1001/jama.2016.14799
shock. Ann Pharmacother. 2023;57(4):375-381. doi:10.1177/ 66. Narum S, Westergren T, Klemp M. Corticosteroids and risk of
10600280221117156 gastrointestinal bleeding: a systematic review and meta-analysis.
52. Tunkel AR, Hartman BJ, Kaplan SL, et al. Practice guidelines for BMJ Open. 2014;4(5):e004587. doi:10.1136/bmjopen-2013-
the management of bacterial meningitis. Clin Infect Dis. 004587
2004;39(9):1267-1284. doi:10.1086/425368 67. Institute TG. Fludrocortisone dose response relationship in
53. de Gans J, van de Beek D. Dexamethasone in adults with bacterial septic shock – FluDReSS. 2021. https://ClinicalTrials.gov/
meningitis. N Engl J Med. 2002;347(20):1549-1556. doi:10.1056/ show/NCT04494789.
NEJMoa021334 68. Wong HR, Sweeney TE, Hart KW, Khatri P, Lindsell CJ.
54. Scarborough M, Gordon SB, Whitty CJM, et al. Corticosteroids Pediatric sepsis endotypes among adults with sepsis. Crit Care
for bacterial meningitis in adults in sub-Saharan Africa. N Engl Med. 2017;45(12):e1289-e1291. doi:10.1097/ccm.00000000000
J Med. 2007;357(24):2441-2450. doi:10.1056/NEJMoa065711 02733
55. Metlay JP, Waterer GW, Long AC, et al. Diagnosis and treatment 69. Scicluna BP, van Vught LA, Zwinderman AH, et al. Classification
of adults with community-acquired pneumonia. An official clini- of patients with sepsis according to blood genomic endotype: a
cal practice guideline of the American Thoracic Society and prospective cohort study. Lancet Respir Med. 2017;5(10):816-
Infectious Diseases Society of America. Am J Respir Crit Care 826. doi:10.1016/s2213-2600(17)30294-1
Med. 2019;200(7):e45-e67. doi:10.1164/rccm.201908-1581ST 70. DeMerle KM, Angus DC, Baillie JK, et al. Sepsis subclasses: a
56. Saleem N, Kulkarni A, Snow TAC, Ambler G, Singer M, framework for development and interpretation. Crit Care Med.
Arulkumaran N. Effect of corticosteroids on mortality and clinical 2021;49(5):748-759. doi:10.1097/ccm.0000000000004842
cure in community-acquired pneumonia: a systematic review, 71. Gordon AC, Mason AJ, Thirunavukkarasu N, et al. Effect of early
meta-analysis, and meta-regression of randomized control trials. vasopressin vs norepinephrine on kidney failure in patients with
Chest. 2023;163(3):484-497. doi:10.1016/j.chest.2022.08.2229 septic shock: the VANISH randomized clinical trial. Jama.
57. Dequin PF, Meziani F, Quenot JP, et al. Hydrocortisone in severe 2016;316(5):509-518. doi:10.1001/jama.2016.10485
community-acquired pneumonia. N Engl J Med. 2023;388(21): 72. Antcliffe DB, Burnham KL, Al-Beidh F, et al. Transcriptomic sig-
1931-1941. doi:10.1056/NEJMoa2215145 natures in sepsis and a differential response to steroids. From the
58. Schäcke H, Döcke W-D, Asadullah K. Mechanisms involved in the VANISH randomized trial. Am J Respir Crit Care Med.
side effects of glucocorticoids. Pharmacol Ther. 2002;96(1):23-43. 2019;199(8):980-986. doi:10.1164/rccm.201807-1419OC
https://doi.org/10.1016/S0163-7258(02)00297-8 73. Fleuriet J, Heming N, Meziani F, et al. Rapid recognition of cor-
59. Puthucheary ZA, Rawal J, McPhail M, et al. Acute skeletal muscle ticosteroid resistant or sensitive sepsis (RECORDS): study proto-
wasting in critical illness. JAMA. 2013;310(15):1591-1600. col for a multicentre, placebo-controlled, biomarker-guided,
doi:10.1001/jama.2013.278481 adaptive Bayesian design basket trial. BMJ Open. 2023;13(3):
60. Schakman O, Gilson H, Thissen JP. Mechanisms of glucocorticoid- e066496. doi:10.1136/bmjopen-2022-066496
induced myopathy. J Endocrinol. 2008;197(1):1-10. doi:10.1677/ 74. Harhay MO, Wagner J, Ratcliffe SJ, et al. Outcomes and statistical
joe-07-0606 power in adult critical care randomized trials. Am J Respir Crit
61. Needham DM, Davidson J, Cohen H, et al. Improving long-term Care Med. 2014;189(12):1469-1478. doi:10.1164/rccm.201401-
outcomes after discharge from intensive care unit: report from a 0056CP