Hydrocortisone is considered the steroid of choice for replacement therapy in patients with acute adrenal crisis and with CIRCI including septic shock with persistent hypotension despite fluid resuscitation [16]. There is no evidence to recommend routine mineralocorticoid (fludrocortisone) replacement for patients in the ICU, unless they have known or suspected primary adrenal insufficiency. Even in Addisonian crisis, hydrocortisone along with saline resuscitation is probably sufficient for initial treatment in the acute setting because hydrocortisone does have some mineralocorticoid activity. A study evaluating hydrocortisone alone vs. hydrocortisone plus fludrocortisone in patients with septic shock failed to show any significant differences in outcomes with the addition of the mineralocorticoid [17].

Steroid administration for treatment of presumed adrenal crisis should never be delayed in order to perform random cortisol and ACTH stimulation testing. Dexamethasone may be useful in such patients because, unlike hydrocortisone, it does not interfere with measurements of serum cortisol. Dexamethasone can cause long-term suppression of the HPA axis, however, and thus it is not generally recommended for ongoing steroid replacement [8].

In suspected acute adrenal crisis, an initial dose of 100 mg hydrocortisone IV followed by 100–200 mg daily is generally adequate [1, 18]. For CIRCI in the setting of severe sepsis and septic shock, 200 mg of hydrocortisone divided into 3–4 intermittent doses daily has been recommended [16]. Hydrocortisone is typically not tapered until patients are no longer requiring vasopressors. Tapering can occur over several days to prevent rebound hypotension and adverse immunologic effects. Discontinuation of steroids after tapering may be contraindicated in patients with adrenal crisis not consistent with CIRCI as they may require ongoing low-dose corticosteroid replacement.

In 2002, a large prospective study demonstrated that 7 days or hydrocortisone and fludrocortisone reduced the 28-day mortality for patients with septic shock [20]. However, the subsequent Corticosteroid Therapy in Septic Shock (CORTICUS) trial failed to replicate any mortality benefit after treatment with hydrocortisone [21]. Patients receiving hydrocortisone did recover from shock more quickly, regardless of results of ACTH stimulation testing, but had a trend toward increased hospital-acquired infections. Furthermore, an observational study by the Surviving Sepsis Campaign showed an increase in hospital mortality in vasopressor-dependent patients receiving low-dose corticosteroid therapy [22].

Systematic reviews and meta-analyses have attempted to resolve the conflicting evidence. Several such studies have indicated that low dose corticosteroid therapy aids in reversal of shock, and thus patients who are vasopressor-dependent despite adequate fluid resuscitation may benefit from corticosteroid therapy [23–27]. However, ACTH stimulation testing has not been shown to accurately discriminate which patients with septic shock will benefit from steroid replacement [8, 16, 26].

Based on this body of evidence, the Surviving Sepsis Campaign no longer recommends routine use of steroid replacement therapy as part of its sepsis care bundles [28]. However, the most recent Surviving Sepsis Guidelines include a Grade 2C (weak) recommendation to consider steroid therapy in patients with severe, persistent, vasopressor-dependent septic shock [16].

Administration of 250 mcg of ACTH is commonly used in ACTH stimulation testing. This “high dose” testing uses supraphysiologic dosing of ACTH, and thus a “low dose” test using only 1 mcg of ACTH has been used as an alternative. The low dose test has better sensitivity, is less expensive, and has been shown in at least one study to better delineate groups of patients with relative adrenal insufficiency in septic shock [4]. However, there is the practical concern of how to accurately dilute the 250 mcg dose, as well as concerns that much of the ACTH dose may bind to injection tubing leading to unknown dose delivery [18]. Furthermore, the high dose ACTH test is better validated, having been used in numerous studies in ICU patients [8, 20, 22, 23, 29] and therefore it remains the favored test in critically ill patients.

Salivary cortisol measurements have been used as a surrogate marker of free cortisol in the diagnosis of adrenal insufficiency [30–32], but salivary cortisol has not been validated as a standalone diagnostic test in adults with suspected CIRCI. A recent study in neurosurgical ICU patients showed a lack of diurnal variation in salivary cortisol production, similar to that seen with serum cortisol, but the authors described challenges in collecting saliva from intubated patients due to decreased salivary flow and dry oral mucosa, medication effects, fluid imbalances, mucosal breakdown, and contamination with blood [33].

Etomidate is an anesthetic agent commonly used during rapid sequence intubation in critically ill patients. Etomidate can cause inhibition of adrenal 11-β-hydroxylase, however, and thus has been implicated in adrenal insufficiency. In the CORTICUS trial, patients who received etomidate were less likely to have a significant increase in cortisol after ACTH, and they had decreased survival at 28 days [21]. In the retrospective CORTICUS study, etomidate administration in critically ill patients was associated with a worse prognosis in patients not treated with steroids [29]. Etomidate has other benefits including a very short duration of action, decreased respiratory depression and less hypotension [34]. The balance of risks and benefits using etomidate for induction of anesthesia during intubation remains unclear in the context of adrenal insufficiency.