Steroid Use in ICU Patients
Primary Recommendation by Clinical Condition
For septic shock unresponsive to fluid resuscitation and moderate-to-high dose vasopressors (>0.1 μg/kg/min norepinephrine equivalent), use IV hydrocortisone 200 mg/day (either as continuous infusion or 50 mg IV every 6 hours) for at least 3 days at full dose. 1, 2 This represents the strongest evidence-based indication for corticosteroids in the ICU setting, with demonstrated mortality benefit in a meta-analysis of 4,268 patients. 2
Condition-Specific Guidelines
Septic Shock (Vasopressor-Dependent)
- Initiate hydrocortisone only when shock remains refractory despite adequate fluid resuscitation AND moderate-to-high dose vasopressors (>0.1 μg/kg/min norepinephrine or equivalent). 1, 2
- Preferred regimen: Hydrocortisone 200 mg/day as continuous IV infusion over 24 hours. 2
- Alternative regimen: Hydrocortisone 50 mg IV bolus every 6 hours if continuous infusion is unavailable. 2
- Duration: Minimum 3 days at full dose, continuing until vasopressors are no longer required. 1, 2
- Enhanced efficacy: Consider adding fludrocortisone to hydrocortisone for combination therapy in septic shock. 3
- Critical caveat: Do NOT use corticosteroids in sepsis without shock—no mortality benefit demonstrated and potential for harm. 1, 2
Acute Respiratory Distress Syndrome (ARDS)
- For early moderate-to-severe ARDS (PaO2/FiO2 <200 within 14 days of onset), use IV methylprednisolone 1 mg/kg/day for ≥14 days. 4, 1, 5
- Methylprednisolone is preferred over hydrocortisone due to greater penetration into lung tissue and longer residence time. 1
- Recent meta-analysis confirms that early initiation (≤72 hours), low-dose, and prolonged therapy (≥7 days) reduces short-term mortality in ARDS. 3
Community-Acquired Pneumonia (CAP)
- For hospitalized patients with CAP, use corticosteroids for 5-7 days at a daily dose <400 mg IV hydrocortisone or equivalent. 4, 1
- Meta-analysis of 13 trials (n=2,005) showed reduced hospital stay (mean difference -2.96 days), decreased need for mechanical ventilation (RR 0.45), and prevented ARDS (RR 0.24). 4
- Subgroup analysis indicates mortality benefit is most pronounced in severe rather than mild pneumonia. 4
Cardiopulmonary Bypass Surgery
- Use corticosteroids in patients undergoing cardiopulmonary bypass surgery (conditional recommendation, moderate quality evidence). 4
- Analysis of 14 trials (n=13,365) found RR of mortality 0.84 (95% CI 0.70-1.01). 4
Critical Dosing Principles
Dose-Response Relationship
- Mortality benefit occurs ONLY with doses <400 mg/day hydrocortisone equivalent for ≥3 days at full dose. 2
- High-dose, short-course regimens (>400 mg/day for <3 days) do NOT improve outcomes and may cause harm. 2
- Survival benefits are dose-dependent: lower doses for longer duration are superior to high-dose, short-course therapy. 2
Agent Selection
- Hydrocortisone is preferred for septic shock because it provides mineralocorticoid activity at physiologic doses. 2
- Network meta-analysis showed hydrocortisone boluses/infusions were more effective than methylprednisolone for shock reversal. 2
- Dexamethasone is NOT recommended for critical illness-related corticosteroid insufficiency. 5
Tapering Protocol
When to Begin Taper
- Start tapering when vasopressors are no longer required, NOT based on arbitrary time points. 1, 2
- Maintain full-dose hydrocortisone for at least 3-5 days before initiating taper. 1
Tapering Duration
- Taper gradually over 6-14 days rather than stopping abruptly to avoid rebound inflammation and hemodynamic deterioration. 1
- One crossover study demonstrated hemodynamic and immunologic rebound effects after abrupt cessation. 1
Monitoring During Taper
- Monitor serum sodium levels for hypernatremia. 1
- Monitor blood glucose for hyperglycemia. 1
- Watch for signs of adrenal insufficiency: hypotension, fever, confusion. 1
Diagnostic Considerations
Do NOT Use ACTH Stimulation Test
- The ACTH stimulation test should NOT be used to identify which patients with septic shock should receive glucocorticoids or to guide tapering decisions. 1, 5
- Clinical criteria (hypotension poorly responsive to vasopressors despite adequate fluid resuscitation) should guide treatment decisions, not laboratory tests. 5, 6
Clinical Diagnosis
- Suspect critical illness-related corticosteroid insufficiency (CIRCI) in hypotensive patients who respond poorly to fluids and vasopressor agents, particularly in sepsis. 5
- Clinical signs include: hypotension refractory to fluid resuscitation, decreased sensitivity to catecholamines, fever, confusion, persistent hypoxia. 1
- Laboratory findings may include: hypoglycemia, hyponatremia, hyperkalemia, metabolic acidosis. 1
Adverse Effects and Monitoring
Expected Adverse Effects
- Hyperglycemia is the most common adverse event (90.9% vs 81.5% in placebo). 2
- Monitor blood glucose regularly and treat hyperglycemia aggressively. 1
- Hypernatremia may occur—monitor serum electrolytes. 1, 2
Reassuring Safety Data
- NO increased risk of secondary infections (RR 1.02,95% CI 0.87-1.20). 2, 3
- NO increased risk of gastrointestinal bleeding. 2, 3
- Infection surveillance is still recommended during treatment as corticosteroids blunt the febrile response. 1
Serious Adverse Effects to Monitor
- Psychiatric effects (agitation, delirium, anxiety). 7
- Muscle weakness with prolonged use. 7
- Fluid retention. 7
Common Pitfalls to Avoid
Timing Errors
- Do NOT delay treatment in suspected adrenal crisis while awaiting diagnostic confirmation. 2
- Do NOT start corticosteroids in sepsis without shock—no benefit and potential harm. 1, 2
Drug Interactions
- Do NOT start other hormone replacements (thyroid, testosterone, estrogen) before corticosteroids—these accelerate cortisol clearance and can precipitate adrenal crisis. 2
- Be aware that etomidate use prior to hydrocortisone initiation may worsen outcomes. 1
Discontinuation Errors
- NEVER stop corticosteroids abruptly after >few days of treatment—this can lead to deterioration from reconstituted inflammatory response. 1, 5
- Reinstitute treatment if signs of sepsis, hypotension, or worsening oxygenation recur after discontinuation. 5
Conditions Where Corticosteroids Are NOT Recommended
Influenza
- Do NOT use corticosteroids in adults with influenza. 4
- Analysis of 13 observational studies (n=1,917) found OR of dying 3.06 (95% CI 1.58-5.92) against corticosteroids. 4
- Increased risk of superinfection documented. 4
Major Trauma
- Corticosteroids are NOT recommended for patients with major trauma. 1
- Analysis of 19 trials showed no significant effect on mortality (RR 1.00,95% CI 0.89-1.13). 1
COVID-19 (Historical Context)
- Early evidence from SARS and MERS showed no clear benefit and potential harm with corticosteroids. 4
- Meta-analysis of SARS treatment reported 25 of 29 studies were inconclusive and 4 demonstrated possible harm. 4
- Multi-center study of 309 ICU patients with MERS showed delayed viral clearance and lack of survival benefit with corticosteroids. 4
Recent Evidence Summary
A 2025 systematic review and meta-analysis of 43 RCTs (n=10,853) demonstrated that corticosteroids reduced short-term mortality in critically ill patients (RR 0.85,95% CI 0.77-0.94). 3 Additional benefits included:
- Reduced ICU length of stay (mean difference -2.02 days). 3
- Reduced hospital length of stay (mean difference -2.66 days). 3
- Reduced duration of mechanical ventilation (mean difference -4.24 days). 3
- Increased ventilator-free days at 28 days (mean difference 2.83 days). 3
- Improved oxygenation index (PaO2/FiO2) by 61.41 mmHg in mechanically ventilated patients. 3
- Reversed shock in sepsis/septic shock (RR 1.20,95% CI 1.06-1.35). 3