Version July 2025
INTRODUCTION —
The role of corticosteroid therapy in patients with septic shock has evolved since the 1990s. Whereas glucocorticoids and mineralocorticoids are distinct classes of hormones, this topic review refers to both glucocorticoids and mineralocorticoids as "corticosteroids."
The rationale and indications for corticosteroid administration are discussed in this topic review. Other aspects of the management of septic shock and role of glucocorticoids in community-acquired pneumonia, coronavirus disease 2019 (COVID-19), and acute respiratory distress syndrome are reviewed separately.
●(See "Evaluation and management of suspected sepsis and septic shock in adults".)
●(See "COVID-19: Management in hospitalized adults", section on 'Dexamethasone and other glucocorticoids'.)
RATIONALE —
Corticosteroids are administered to patients with septic shock to counteract absolute or relative adrenal insufficiency and to suppress excessive inflammation that may contribute to shock. This may help restore balance to the altered hypothalamic-pituitary-adrenal axis and inflammatory response with the goal of improving clinically meaningful outcomes, such as mortality.
ADRENAL INSUFFICIENCY IN SEPSIS
Relative adrenal insufficiency — Absolute adrenal insufficiency is rare among critically ill patients, with an incidence estimated to be ≤3 percent [1]. However, "relative adrenal insufficiency" or "critical illness-related corticosteroid insufficiency" (CIRCI) may be common among critically ill patients (ie, adrenal cortisol production is insufficient to meet the body's demands) [2].
There is no agreement about diagnostic criteria of CIRCI. In addition, there is considerable disagreement over what qualifies as a "normal" or "appropriate" cortisol level in septic shock, what dose of adrenocorticotropic hormone should be used for stimulation testing, and what constitutes an adequate response. Without a clear definition and reliable assays in critically ill patients, whether CIRCI is a true diagnostic entity remains speculative.
Mechanisms of adrenal insufficiency in sepsis — Critical illness affects cortisol level and function through several mechanisms including hypothalamic-pituitary-adrenal (HPA) axis activation (resulting in increased levels of circulating cortisol), HPA impairment (resulting in adrenocortical hyporesponsiveness), and glucocorticoid resistance.
●Activation of the HPA axis – In patients who are critically ill, serum cortisol rises, reaching levels as high as 40 to 50 mcg/dL (figure 1) [3-7]. Cortisol metabolism and function may also be altered in sepsis resulting in reduced cortisol breakdown [7], reduced binding of cortisol to albumin and cortisol binding globulin [8,9], increased glucocorticoid receptor affinity for cortisol, and peripheral conversion of precursors to cortisol [10,11].
●Impairment of the HPA axis – Among critically ill patients, several factors are known to impair the HPA axis including head injury, central nervous system depressants, pituitary infarction, adrenal hemorrhage, infections, malignancy, previous glucocorticoid therapy, and several drugs (eg, ketoconazole, phenytoin, and etomidate) [12-14].
●Glucocorticoid resistance – Glucocorticoid resistance may also play a role. One study reported that compared with healthy controls, patients with sepsis demonstrated higher expression levels of the beta-isoform of the glucocorticoid receptor, an isoform associated with glucocorticoid resistance [15]. However, functional differences were not observed. Other mechanisms of glucocorticoid resistance are discussed separately. (See "Glucocorticoid resistance in asthma", section on 'Potential mechanisms of glucocorticoid resistance'.)
ASSESSING ADRENAL RESERVE (NOT TYPICALLY PERFORMED) —
In general, most clinicians do not rely on laboratory testing to select glucocorticoid replacement therapy in patients with septic shock. This is because laboratory assays of plasma cortisol concentration and response to adrenocorticotropic hormone (ACTH) stimulation are unreliable in critically ill patients. In addition, in major randomized trials, baseline cortisol levels and the ACTH stimulation test have failed to consistently identify patients with septic shock who benefit from glucocorticoid use. (See 'Efficacy for refractory septic shock' below.)
With this caveat in mind, international guidelines for clinicians from the Society of Critical Care Medicine (SCCM) and European Society of Intensive Care Medicine (ESICM) [2] endorse the following as indicators of likely adrenal insufficiency in critically ill patients:
●A change in baseline cortisol of <9 mcg/dL 60 minutes after administration of cosyntropin (250 mcg; ie, high-dose ACTH stimulation).
●A random plasma cortisol of ≤10 mcg/dL.
Studies using both low-dose (1 mcg cosyntropin) and high-dose ACTH stimulation (250 mcg cosyntropin) have yielded variable results in critically ill patients [16-21]. If clinicians wish to perform an ACTH stimulation test, we use the high-dose test as it is easier to perform [2]; in addition, commercial ampoules contain preprepared 250 mcg of cosyntropin while the low-dose test often requires some preparation at the bedside. (See "Diagnosis of adrenal insufficiency in adults", section on 'ACTH stimulation tests'.)
ACTH stimulation testing for non-critically ill patients is discussed separately. (See "Diagnosis of adrenal insufficiency in adults", section on 'ACTH stimulation tests'.)
Data supporting the unreliability of laboratory tests for adrenal insufficiency in critically ill patients include the following:
●Variability in random serum cortisol levels – In patients with septic shock, total serum cortisol levels vary widely [6,16,17,22-25] and do not correlate with mortality [16,18,24,26-32].
●Limited ability to measure free cortisol – In critically ill patients, there is a shift from inactive protein-bound cortisol to physiologically active free cortisol [33,34]. However, standard assays for plasma cortisol measure total (free + bound) plasma cortisol and free cortisol assays are not available at most clinical centers [35].
●Poor reliability of ACTH stimulation tests – ACTH stimulation tests are unreliable in critically ill patients for several reasons:
•Some critically ill individuals have spontaneous increases in their serum cortisol of ≥9 mcg/dL without cosyntropin stimulation [36].
•ACTH stimulation tests may give inconsistent results in the same individuals if performed on more than one occasion [16,37].
•In septic shock, standard immunoassays appear to correlate poorly with the reference standard (mass spectrometry) [38].
•Etomidate, if used for intubating patients with septic shock, can interfere with the results of ACTH stimulation since it suppresses the hypothalamic-pituitary-adrenal axis [39,40]. However, a randomized trial found that while etomidate impaired the response to ACTH (compared with ketamine), it was not associated with worse clinical outcomes [39].
PATIENT SELECTION —
When considering patients with septic shock for corticosteroid therapy, we generally use the approach outlined below.
Indications — Adult patients with septic shock in whom corticosteroid therapy is administered include the following:
●Patients with septic shock from conditions that benefit from glucocorticoid therapy – We administer systemic glucocorticoid therapy to adult patients with septic shock from conditions that benefit from such therapy. This includes the following:
•A steroid-responsive process (eg, acute eosinophilic pneumonia, organizing pneumonia, hemophagocytic lymphohistiocytosis) – Agent selection and dosing for these indications are provided separately. (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Cryptogenic organizing pneumonia", section on 'Treatment' and "Treatment and prognosis of hemophagocytic lymphohistiocytosis", section on 'Acutely ill or deteriorating patients'.)
•Acute respiratory distress syndrome (ARDS) refractory to standard therapies – Patient and agent selection and dosing for this indication are provided separately. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults".)
•Severe COVID-19 – Patient and agent selection and dosing for this indication are provided separately. (See "COVID-19: Management in hospitalized adults", section on 'Dexamethasone and other glucocorticoids'.)
•Severe community-acquired pneumonia (CAP) – Patient and agent selection and dosing for this indication are provided separately. (See "Treatment of community-acquired pneumonia in adults who require hospitalization", section on 'Adjunctive glucocorticoids'.)
●Refractory septic shock – We suggest corticosteroid therapy in adult patients with refractory septic shock (defined as a systolic blood pressure <90 mmHg for more than one hour following both adequate fluid resuscitation and vasopressor administration). This approach is based upon data suggesting faster resolution of shock and a possible mortality benefit that may depend upon the regimen used and patient phenotype. Supportive data, agent selection, and dosing are provided below. (See 'Efficacy for refractory septic shock' below and 'Administration' below.)
Avoidance
●For adult patients with septic shock, we do not routinely use corticosteroid therapy as part of the initial therapeutic regimen. Patients should be monitored and subsequently evaluated for their response to fluid resuscitation and vasopressor administration.
●We also generally avoid glucocorticoids in patients with severe influenza since harm may be associated with glucocorticoids. (See "Seasonal influenza in nonpregnant adults: Treatment", section on 'Investigational approaches'.)
Efficacy for refractory septic shock — Trials in critically ill patients with sepsis prior to the 2000s were conflicting [41-46] and prompted several larger randomized trials, which are discussed in the sections below. Data have consistently shown that corticosteroid therapy leads to faster resolution of shock. In contrast, a mortality benefit is less consistent and may depend upon the regimen used and individual patient or disease characteristics. Limited data suggest that administration within 12 hours of presentation may provide greater benefit compared with administration >12 hours after presentation [47].
Meta-analyses — Several meta-analyses have demonstrated efficacy of corticosteroids in patients with refractory septic shock [48-51].
●A 2024 network meta-analysis of 17 trials compared hydrocortisone (a glucocorticoid) plus fludrocortisone (a mineralocorticoid) with hydrocortisone alone using aggregate data (as opposed to individual patient data) [48]. All-cause mortality at last follow-up was lowest with fludrocortisone plus hydrocortisone (relative risk [RR] 0.85, 95% CrI 0.72-0.99), followed by hydrocortisone alone (RR 0.97, 95% CrI 0.87-1.07). Fludrocortisone plus hydrocortisone was associated with a 12 percent reduction in the risk of all-cause mortality compared with hydrocortisone alone (RR 0.88, 95% CrI 0.74-1.03; 3.8 percent absolute risk reduction, 95% CrI -8.6 to +1.1 percent). However, the evidence was mostly indirect and only two trials used a direct comparison. Safety outcomes were similar between treatment regimens.
●A 2023 meta-analysis analyzed individual patient data from 17 trials (14 of which overlapped with the above). Ninety-day mortality was similar when all hydrocortisone regimens were compared with placebo (RR 0.93, 95% CI 0.82-1.04). Subgroup analysis reported a reduction in 90-day mortality when a hydrocortisone plus fludrocortisone regimen was compared with placebo (RR 0.88, 95% CI 0.81-0.97) [51]. However, mortality was similar when the hydrocortisone plus fludrocortisone regimen was compared with hydrocortisone alone (RR 0.93, 95% CI 0.83-1.03). Hydrocortisone-containing regimens were associated with more vasopressor-free days when compared with placebo (mean difference 1.24 days, 95% CI 0.74-1.73). Safety outcomes were similar except that hydrocortisone may be associated with an increased risk of hypernatremia (RR 2.01, 95% CI 1.56-2.6) and muscle weakness (RR 1.73, 95% CI 1.49-1.99). However, among the 17 eligible trials, only 10 were at low risk of bias.
●The Society of Critical Care Medicine (SCCM) also performed a meta-analysis on 46 randomized trials in several patient groups including ARDS and CAP, which supported their recommendation for the administration of glucocorticoids in patients with refractory shock [52,53].
Hydrocortisone plus fludrocortisone — Two of three major randomized trials demonstrated faster resolution of shock together with a mortality benefit from the administration of both hydrocortisone (a glucocorticoid) and fludrocortisone (a mineralocorticoid) when compared with placebo [54,55]. The third trial failed to demonstrate a mortality benefit when hydrocortisone plus fludrocortisone was compared with hydrocortisone alone but was underpowered for this comparison [56].
●French trial – In this 2002 trial of 300 patients with vasopressor-dependent shock, hydrocortisone (50 mg intravenously every six hours) plus fludrocortisone (50 mcg enterally once a day) started within eight hours and continued for seven days resulted in decreased 28-day mortality (55 versus 61 percent) and faster shock reversal (seven versus nine days) compared with placebo [54]. The trial was criticized for its high placebo-group mortality [57-59].
●Activated Protein C and Corticosteroids for Human Septic Shock (APROCCHHS) – The same group published another trial in 2018. In 1241 patients with severe vasopressor-dependent septic shock (mixed surgical and medical patients), a seven-day course of hydrocortisone/fludrocortisone administered at the same doses listed above decreased 90- and 180-day mortality (43 versus 49 percent [RR 0.88, 95% CI 0.78-0.99] and 47 versus 53 percent, respectively) [55]. Vasopressor-free days were increased compared with placebo (17 versus 15 days). There was no increase in the rates of superinfection or neurologic sequelae, but corticosteroids resulted in an increase in the rate of hyperglycemia (89 versus 83 percent). The mortality benefit was maintained in a prespecified subgroup analysis for patients with sepsis from CAP [60]. (See "Treatment of community-acquired pneumonia in adults who require hospitalization", section on 'Adjunctive glucocorticoids'.)
●Corticosteroid Treatment and Intensive Insulin Therapy for Septic Shock in Adults (COIITSS) – In this trial of 509 patients with septic shock, hydrocortisone plus fludrocortisone resulted in a 3 percent absolute risk reduction in the risk of death compared with hydrocortisone alone, but the difference was not significant [56]. However, the trial was underpowered to detect a clinically important effect. A similar but significant absolute risk reduction was seen in a large observational database study [49].
Differences between trials that used hydrocortisone plus fludrocortisone (demonstrating a mortality benefit) [54-56] and those that used hydrocortisone alone (none of which showed a mortality benefit) (see 'Hydrocortisone alone' below) may be explained by the following:
●Addition of fludrocortisone – Traditionally, the term "mineralocorticoid activity" was used to describe corticosteroid activity regulating electrolyte balance [61]. However, while published "mineralocorticoid equivalencies" [61] based on sodium retaining properties alone suggest a 1:125 potency ratio for hydrocortisone compared with fludrocortisone, the relative daily dose of hydrocortisone to fludrocortisone in the two trials above showing benefit of combination therapy greatly exceeds this ratio (hydrocortisone 200 mg versus fludrocortisone 0.05 mg). This suggests that other mechanisms contribute to the benefits of fludrocortisone in this setting. For example, fludrocortisone has nonkidney effects [62,63] including inhibition of interleukin-1 production [64], activation of the innate immune system [62,65], a neural antiapoptotic role [66], and increased clearance of excess alveolar fluid by pulmonary epithelial cells [62,67,68]. Additionally, exploratory data found that mineralocorticoid synthesis was substantially lower in sepsis nonsurvivors than survivors whereas glucocorticoid synthesis was not statistically different [69].
●Sicker patients – Both the initial French trial and APROCCHHS were comprised of a sicker group of patients as evidenced by the high simplified acute physiology II scores (SAPS II), greater need for higher doses of vasopressors at the time of enrollment, and higher mortality risk. (See "Predictive scoring systems in the intensive care unit".)
●Different sepsis sources – Sources of infections may have also been different. For example, APROCCHHS had fewer patients with abdominal infection following surgery and had more patients with lung infections than trials that did not suggest mortality benefit. However, a meta-analysis reported that the source of infection did not influence the effectiveness of hydrocortisone [51]. Further trials are needed to determine whether source of infection influences treatment efficacy.
Hydrocortisone alone — Several major randomized trials demonstrated faster resolution of shock without a mortality benefit from the administration of hydrocortisone:
●Corticosteroid Therapy of Septic Shock (CORTICUS) – In this 2008 trial of 499 patients with septic shock (regardless of vasopressor dependency), compared with placebo, hydrocortisone administration (200 mg daily) resulted in a faster reversal of shock (3.3 versus 5.8 days), but did not improve 28-day mortality (35 versus 32 percent in the placebo group) [70]. Although new infection rates were higher in the hydrocortisone group, they did not reach statistical significance (odds ratio 1.27, 95% CI 0.96-1.68). The trial was criticized for the lower-than-expected placebo-group mortality (32 percent versus the anticipated 50 percent).
●Effect of Early Vasopressin versus Norepinephrine on Kidney Failure in Patients with Septic Shock (VANISH) – In this 2016 trial that compared vasopressin with norepinephrine in patients with septic shock, when compared with placebo, the addition of hydrocortisone to either vasopressor did not result in a mortality benefit (28-day mortality 33 versus 29 percent for vasopressin group and 29 versus 26 percent for norepinephrine group) or have an effect on the rate of kidney failure [71].
A subsequent post-hoc analysis demonstrated heterogeneity of the treatment effect [72]. Patients' transcriptomic patterns were associated with different responses to hydrocortisone, resulting in a net-neutral trial (ie, some patients may have benefitted whereas others may have been harmed). However, these data are preliminary and biomarker profiles are not yet used in routine clinical practice.
●Adjunctive Corticosteroid Treatment in Critically Ill Patients with Septic Shock (ADRENAL) – In this 2018 trial of 3800 patients (medical and surgical) with vasopressor-dependent septic shock, hydrocortisone resulted in a faster resolution of shock (three versus four days) and shorter duration of initial mechanical ventilation (six versus seven days) compared with placebo [73]. However, hydrocortisone administration did not improve 28- or 90-day mortality, overall number of mechanical ventilation-free days, rate of recurrent shock, or rate of kidney replacement therapy. Higher rates of hyperglycemia and hypernatremia were observed in the hydrocortisone group. A post-hoc analysis of this trial that applied updated Sepsis-3 or Sepsis-2 inclusion criteria to the study population still reported no mortality benefit from hydrocortisone [74].
●Hydrocortisone for Prevention of Septic Shock (HYPRESS) – This 2016 trial of 353 patients who had severe sepsis but not septic shock reported similar mortality rates and rate of progression to septic shock [75] (table 1 and table 2). Hydrocortisone was associated with an increased rate of hyperglycemia (91 versus 82 percent) and a nonsignificant increase in the rate of infections (22 versus 17 percent) and muscle weakness (31 versus 24 percent). Criticisms of this trial included the exclusion of patients with septic shock.
Specific populations (pneumonia, COVID-19, ARDS) — Data on the benefits of glucocorticoids in other disorders that can be associated with septic shock are discussed elsewhere:
•Septic shock in association with a steroid-responsive process (eg, acute eosinophilic pneumonia, organizing pneumonia, hemophagocytic lymphohistiocytosis). (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Cryptogenic organizing pneumonia", section on 'Treatment' and "Treatment and prognosis of hemophagocytic lymphohistiocytosis", section on 'Acutely ill or deteriorating patients'.)
•ARDS associated with COVID-19. (See "COVID-19: Management in hospitalized adults", section on 'Dexamethasone and other glucocorticoids'.)
•ARDS associated with CAP. (See "Treatment of community-acquired pneumonia in adults who require hospitalization", section on 'Adjunctive glucocorticoids'.)
Subgroup variation — Preliminary evidence suggests that, based on biomarker profiles, some patient subgroups may respond favorably to corticosteroid therapy [76-78] whereas others may not benefit or may experience harm [72,77,78]. Biomarkers are not currently used in routine clinical practice to select patients who should receive corticosteroid therapy [79].
ADMINISTRATION
Agent selection — For patients with septic shock in whom corticosteroids are indicated (see 'Patient selection' above), we suggest hydrocortisone (<400 mg intravenously per day in divided doses) as the glucocorticoid of choice. When combination therapy is desired (our preference), we add fludrocortisone (0.05 to 0.1 mg orally or via gastric tube once daily) based upon data that suggest a possible mortality benefit with this combination when compared with hydrocortisone alone (see 'Meta-analyses' above and 'Hydrocortisone plus fludrocortisone' above). Our approach is consistent with that of others [80]. However, practice still varies among clinicians and many still administer hydrocortisone alone.
If hydrocortisone cannot be administered (eg, shortage) or the patient requires another glucocorticoid for a different indication (eg, dexamethasone for cerebral edema or COVID-19-related acute respiratory distress syndrome [ARDS]), the clinician should compare the anti-inflammatory potency of the selected agent with that of hydrocortisone (table 3) and provide equivalent dosing, if feasible, without compromising the dosing for another indication. The assistance of a pharmacist may be useful in this regard. Our practice is to add fludrocortisone in septic shock to other glucocorticoid regimens, which will be the focus of future clinical trials.
Agent selection and dosing for refractory septic shock are discussed here while agent selection and dosing for other conditions associated with septic shock are provided separately:
●Septic shock in association with a steroid-responsive process (eg, acute eosinophilic pneumonia, organizing pneumonia, hemophagocytic lymphohistiocytosis). (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Cryptogenic organizing pneumonia", section on 'Treatment' and "Treatment and prognosis of hemophagocytic lymphohistiocytosis", section on 'Acutely ill or deteriorating patients'.)
●ARDS associated with COVID-19. (See "COVID-19: Management in hospitalized adults", section on 'Dexamethasone and other glucocorticoids'.)
●ARDS associated with community-acquired pneumonia. (See "Treatment of community-acquired pneumonia in adults who require hospitalization", section on 'Adjunctive glucocorticoids'.)
Dosing
Glucocorticoid — Dosing varies with the agent used:
●Hydrocortisone – We typically administer 200 to 300 mg per day of intravenous hydrocortisone in divided doses (50 mg every six hours or 100 mg every eight hours). This is consistent with guidelines set out by the Society of Critical Care Medicine (SCCM) and European Society of Intensive Care Medicine (ESICM) who recommend <400 mg/day of hydrocortisone [52].
Most studies evaluating the effect of corticosteroids in septic shock used hydrocortisone as the preferred glucocorticoid in divided doses rather than an infusion. Although one small prospective study showed less variability in blood glucose levels with continuous hydrocortisone infusions, data suggest no significant difference in 90-day mortality between continuous versus bolus administration of hydrocortisone [51].
●Methylprednisone – Dosing for methylprednisolone is less well studied, but one suggested regimen is 40 to 60 mg intravenously daily in divided doses for a similar time period. This dosing regimen has an anti-inflammatory effect that closely approximates that of the hydrocortisone regimen.
Mineralocorticoid — Fludrocortisone is administered as 0.05 or 0.1 mg once daily for five to seven days (in combination with intravenous hydrocortisone or equivalent) [54,55,81]. Although enteral absorption may be questioned in patients with shock, limited data suggest that is not the case [82,83]. Preliminary data support the higher dosing of 0.1 mg [83], but practice varies regarding its use. Noteworthy is that, in some countries, formulations are only available as 0.1 mg tablets (but can be cut to obtain the 0.05 mg dose).
Duration — Regimens vary; our approach is the following:
●Duration – We typically administer hydrocortisone for five to seven days. This practice is consistent with SCCM/ESICM guidelines [52] and that used in randomized trials. (See 'Efficacy for refractory septic shock' above.)
●Tapering – Our practice is to prescribe hydrocortisone and fludrocortisone for a fixed duration of five to seven days without tapering. This approach is based upon the two randomized trials that suggested mortality benefit from regimens that did not include tapering [54,55]. In addition, a meta-analysis reported similar mortality when corticosteroid regimens with tapering were compared with abrupt discontinuation. Proponents of tapering base the practice on preventing rebound hypotension, but data have shown mixed effects with respect to this outcome [84,85]. If shock recurs after cessation without taper, we consider restarting corticosteroids and initiate further evaluation for residual shock [86].
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sepsis in children and adults".)
SUMMARY AND RECOMMENDATIONS
●Critical illness-related adrenal insufficiency
•Scope – Absolute adrenal insufficiency is rare in critically ill patients. Relative adrenal insufficiency (ie, critical illness-related corticosteroid insufficiency) is more common, although a clear definition is lacking. (See 'Relative adrenal insufficiency' above.)
•Mechanism – In patients with septic shock, several mechanisms lead to adrenal insufficiency including increased levels of circulating cortisol, adrenocortical hyporesponsiveness, and glucocorticoid resistance. (See 'Mechanisms of adrenal insufficiency in sepsis' above.)
•Laboratory testing – We do not use laboratory testing to select corticosteroid therapy in patients with septic shock because measurement of plasma cortisol and adrenocorticotropic hormone stimulation testing are unreliable in this population and cannot predict those who may benefit. (See 'Assessing adrenal reserve (not typically performed)' above.)
●Our approach – Adult patients with septic shock in whom corticosteroid therapy is administered include the following:
•Patients with septic shock from conditions that benefit from glucocorticoid therapy – These indications, agent selection, and dosing are provided separately:
-A steroid-responsive process (eg, acute eosinophilic pneumonia, organizing pneumonia, hemophagocytic lymphohistiocytosis) – (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Cryptogenic organizing pneumonia", section on 'Treatment' and "Treatment and prognosis of hemophagocytic lymphohistiocytosis", section on 'Acutely ill or deteriorating patients'.)
-Acute respiratory distress syndrome refractory to standard therapies – (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults".)
-Severe COVID-19 – (See "COVID-19: Management in hospitalized adults", section on 'Dexamethasone and other glucocorticoids'.)
-Severe community-acquired pneumonia – (See "Treatment of community-acquired pneumonia in adults who require hospitalization", section on 'Adjunctive glucocorticoids'.)
•Refractory septic shock – In adult patients with refractory septic shock, we suggest corticosteroid administration (Grade 2B). Refractory shock is defined as a systolic blood pressure <90 mmHg for more than one hour following both adequate fluid resuscitation and vasopressor administration. This approach is based upon data showing faster resolution of shock and a possible mortality benefit that may depend upon the regimen used and patient phenotype.
●Corticosteroid regimen – Our approach mimics regimens used in studies that show benefit:
•Agent selection and dosing – When indicated, we suggest intravenous hydrocortisone plus enteral fludrocortisone rather than other corticosteroid regimens (Grade 2C). Typical dosing of hydrocortisone is 200 to 300 mg intravenously per day in divided doses and of fludrocortisone is 0.05 to 0.1 mg enterally once daily. We use this regimen based upon limited data that suggest a possible mortality benefit with this combination. (See 'Agent selection' above and 'Dosing' above.)
•Duration – Corticosteroids are typically administered for five to seven days without taper. If shock recurs after cessation, corticosteroids may be restarted and patients re-evaluated for residual shock. (See 'Duration' above.)
ACKNOWLEDGMENT —
The UpToDate editorial staff acknowledges Jordi Mancebo, MD, who contributed to earlier versions of this topic review.
