Skip to main content

Foundational Trials

ADRENAL

Image: Shutterstock / luchschenF

Paper

Sections

Publication  |  Context & Rationale  |  Design & Methods  |  Key Results  |  Internal Validity  |  External Validity  |  Strengths & Limitations  |  Interpretation & Why It Matters  |  Controversies & Subsequent Evidence  |  Summary  |  Further Reading  |  Overall Takeaway

Publication

  • Title

    Adjunctive Glucocorticoid Therapy in Patients with Septic Shock​

  • Acronym

    ADRENAL

  • Year

    2018

  • Journal

    New England Journal of Medicine

  • Citation

    Venkatesh B, Finfer S, Cohen J, et al. for the ADRENAL Trial Investigators and ANZICS Clinical Trials Group. N Engl J Med 2018;378(9):797-808

Context & Rationale

  • Background
    • Low-dose corticosteroids have been used as adjunctive therapy in septic shock for decades, but their impact on survival was unclear. Early trials (e.g. Annane 2002) suggested mortality benefits in some patients, whereas later studies (e.g. CORTICUS 2008) found no survival improvement​.
    • By 2016, international guidelines gave only a weak recommendation for hydrocortisone in septic shock refractory to fluids and vasopressors, reflecting low-quality evidence.
    • Meta-analyses up to 2017 reached conflicting conclusions on whether steroids reduce short-term mortality.
    • In short, clinicians lacked a definitive answer on whether steroids truly improve outcomes or merely speed up shock resolution.
  • Research Question
    • The ADRENAL trial was designed to fill this evidence gap. The investigators asked: In adults with vasopressor-dependent septic shock on mechanical ventilation, does intravenous hydrocortisone (200 mg/day) compared to placebo improve 90-day all-cause mortality?​
    • The hypothesis was that adding low-dose hydrocortisone would increase survival in septic shock, based on its potential to reverse shock and modulate inflammatory injury.
  • Why This Matters
    • Septic shock carries high mortality (30–50% historically), and corticosteroids are widely available, inexpensive, and frequently used despite uncertainty.
    • A large, rigorous trial could definitively guide practice – confirming if steroids save lives or if their benefits are limited to faster hemodynamic stabilization.
    • Given the prevalence of sepsis, even a modest mortality benefit would impact many patients, whereas evidence of no benefit would spare patients unnecessary drug exposure and side effects.
    • The ADRENAL trial’s results were therefore highly anticipated to resolve a long-standing controversy in critical care.

Design & Methodology

Trial Design

  • Design
    • ADRENAL was a multi-center, randomized, double-blind, placebo-controlled trial​.
    • Randomization was centralized (web-based) and stratified by site and admission type (medical vs surgical)​, ensuring allocation concealment. The trial was investigator-initiated (funded by academic/government grants) and conducted in 69 ICUs across five countries (Australia, New Zealand, UK, Saudi Arabia, Denmark)​.
    • Enrollment ran from March 2013 to April 2017​

Population

  • Inclusion Criteria

    Adult critically ill patients with septic shock were eligible.

    • documented or strongly suspected infection,
    • ≥2 SIRS criteria,
    • vasopressor dependence for ≥4 hours to maintain BP (SBP >90 or MAP >60 mmHg) despite fluid resuscitation
    • mechanical ventilation (invasive or non-invasive)
    • patients had to be enrolled within 24 hours of meeting shock criteria.
  • Exclusion Criteria
    • prolonged shock (>24h since eligibility)​,
    • an expectation by treating clinicians to use systemic corticosteroids for another indication (e.g. chronic steroid therapy, adrenal insufficiency)​,
    • recent treatment with etomidate or amphotericin B (drugs that affect adrenal function)​,
    • certain infections where steroids could be harmful (cerebral malaria or Strongyloides hyperinfection risk)​
    • patients for whom death was imminent or likely within 90 days due to underlying disease (limiting the trial’s focus to those with potentially reversible illness)​
    • pregnant patients were also excluded (per protocol, though not explicitly listed above).
    • In total 3,800 patients were randomized, of whom 3,658 were included in the primary analysis (142 patients [3.7%] withdrew consent or were lost to follow-up)​
    • Baseline characteristics were well-matched between groups: e.g., mean age ~62 years, ~60% male, ~31% surgical admissions, median APACHE II ~23–24​, with nearly all patients on norepinephrine and mechanical ventilation at enrollment​

Intervention

  • Hydrocortisone

    200 mg/day administered via continuous IV infusion for up to 7 days or until ICU discharge (whichever occurred first), or death.

Control

  • Placebo

    Placebo continuous infusion, matched in appearance and timing to maintain blinding.

Statistical Plan

  • Power Calculation

    The investigators hypothesized a baseline 90-day mortality of ~37% in septic shock. They aimed to detect a 5% absolute mortality reduction (from 37% to 32%) with 90% power at a 5% significance level, requiring approximately 3,800 patients.

  • Analysis

    Primary analyses followed the intention-to-treat principle.

Other

  • Blinding

    Double-blind: Patients, treating clinicians, investigators, and study personnel remained unaware of treatment allocation, minimizing performance and detection bias.

  • Follow Up

    Primary outcome assessed at 90 days post-randomization.

Key Results

Hydrocortisone vs Placebo

  • Early Stopping

    This trial was not stopped early; it completed enrollment as planned.

Primary Outcome

  • 90-day mortality
    • 27.9% (511/1,840) vs 28.8% (526/1,858)
    • Absolute Difference: -0.9%
    • Relative Risk 0.97 (95% CI: 0.90–1.03)
    • P = 0.50

Secondary Outcomes

  • Median time to shock reversal
    • 3 days vs 4 days
    • P <0.001
    • Faster shock reversal with hydrocortisone
  • Days alive and free of vasopressors
    • 17 (IQR 11–20) vs 15 (IQR 9–20)
    • P = 0.003
  • Serious adverse events

    Overall adverse event profile was comparable

Notes

  • No statistically significant difference in 90-day all-cause mortality.
  • Hydrocortisone was associated with faster resolution of shock and more vasopressor-free days.
  • Serious adverse event rates were not significantly different between groups.

Internal Validity

  • Randomisation & Allocation
    • Central randomization with robust allocation concealment.
    • Large sample size with well-executed stratification ensured balance across treatment arms.
  • Performance/Detection Bias
    • Double-blinding reduced risk of bias in both intervention delivery and outcome assessment.
  • Protocol Adherence
    • Protocol guidelines for continuous infusion of hydrocortisone (or placebo) were followed consistently across sites.
    • Compliance data indicate that the intervention and control arms were well separated (i.e., near-complete adherence to assigned treatment).
  • Outcome Assessment
    • Primary endpoint (90-day all-cause mortality) is a robust, objective measure.
    • Secondary outcomes (time to shock reversal) were well-defined and systematically recorded.
  • Statistical Rigor
    • Prespecified statistical analysis plan was followed.
    • Sufficient sample size was accrued, satisfying the initial power calculation.
  • Separation of the Variable of Interest
    • Actual delivered hydrocortisone: 200 mg/day (infusion).
    • Placebo group: 0 mg, but same infusion schedule for blinding.
  • Timing
    • Intervention (hydrocortisone or placebo) was commenced within 24 hours of meeting septic shock criteria.
    • This aligns with prior suggestions that early intervention may be essential in sepsis management.
  • Drop Outs or Exclusions
    • Minimal loss to follow-up, with high retention at 90 days.
    • Post-randomization exclusions were rare and did not appreciably impact the final analysis.
  • Baseline Characteristics
    • Comparable between groups (age, illness severity, site of infection).
    • Patients had severe shock requiring vasopressors plus mechanical ventilation, matching the high-severity phenotype for which steroids have been hypothesized to be beneficial.
  • Heterogeneity
    • Centers spanned multiple countries
    • No major indications that treatment effect varied significantly by region or baseline characteristics.
  • Dose
    • 200 mg/day via continuous infusion for a maximum of 7 days.
    • This dose is often regarded as the standard “stress-dose” hydrocortisone regimen in septic shock trials.
  • Key Delivery Aspects
    • Patients were enrolled relatively early in their ICU course; the control group received standard sepsis care (including vasopressors, fluids, and antimicrobials).
    • The protocol mirrored typical clinical practice, lending realism to the management approaches.
  • Conclusion

    Overall, internal validity appears strong, reflecting rigorous randomization, blinding, consistent protocol adherence, and sound statistical analysis.

External Validity

  • Population Representativeness
    • Enrolled patients were typical of ICU populations worldwide with severe septic shock.
    • Few absolute exclusion criteria, so the results apply broadly to critically ill, vasopressor-dependent, mechanically ventilated adults with sepsis.
  • Applicability
    • The findings are likely generalizable to a range of hospital settings in high-income countries.
    • While resource-limited settings were underrepresented, the pathophysiology of septic shock is largely consistent across regions.
  • Conclusion
    • ADRENAL’s findings are highly generalizable to most ICU settings where standard sepsis care (fluids, vasopressors, antibiotics) is readily available.

Strengths & Limitations

Strengths

  • Large RCT

    Large, multicenter, double-blind RCT with a representative ICU population.

  • Robust Methodology

    Robust methodology with high protocol adherence.

  • Clinically important endpoints

    Clinically important endpoints (90-day mortality, shock resolution).

Limitations

  • High resource setting

    Conducted primarily in well-resourced centers; applicability to low-resource settings is less clear.

  • No cost analysis

    No formal cost-effectiveness analysis.

  • No Effect on Primary Outcome

    Although clinically relevant secondary benefits (faster shock reversal, reduced vasopressor use) were noted, these did not translate into a mortality advantage at 90 days.

Interpretation / Why This Matters

  • Important Primary Outcome

    The ADRENAL trial suggests that while hydrocortisone may accelerate shock resolution and reduce vasopressor requirements, it does not reduce all-cause mortality at 90 days in this broad population of septic shock patients.

  • Highly Clinically Relevant

    This is an important finding for clinicians who regularly weigh the risks and benefits of steroid therapy.

  • Resource Implications

    Faster resolution of shock is a meaningful patient-centered outcome, potentially improving ICU workflow and resource utilization, even if it does not improve long-term survival.

Controversies & Subsequent Evidence

  • Subgroup Effects?

    The lack of mortality benefit spurred discussion as to whether certain subgroups (e.g., refractory shock or specific infection etiologies) might benefit more distinctly.

  • APROCCHSS Trial Comparison

    Some commentators have contrasted ADRENAL with the contemporaneous APROCCHSS trial, which did suggest a mortality benefit using hydrocortisone plus fludrocortisone. Differences in study design and patient populations may account for discrepancies.

  • Meta Analyses

    Subsequent meta-analyses confirm that steroids can hasten shock reversal but consistently show marginal or no mortality advantage, fueling ongoing debate about the ideal patient subgroup, timing, and steroid regimen.

Summary

  • Large

    Large, international, double-blind RCT of hydrocortisone in septic shock.

  • No mortality effect

    No significant reduction in 90-day mortality with hydrocortisone vs. placebo.

  • Faster shock reversal

    Faster shock reversal and more vasopressor-free days in the hydrocortisone group.

  • Robust

    Strong internal validity and broad applicability to ICU care.

  • Impactful

    Results shape modern understanding of adjunctive steroid use in septic shock.

Conclusion

  • No mortality benefit

    A continuous infusion of hydrocortisone, at a dose of 200 mg/day, does not reduce mortality in patients with septic shock

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Overall Takeaway

ADRENAL is a landmark trial in critical care which clarified the role of adjunctive hydrocortisone in septic shock. Its rigorous design and large sample size showed that while steroids speed shock resolution, they do not substantially affect survival at 90 days. This informs nuanced steroid use in septic shock management, reinforcing the idea that hydrocortisone may be a useful adjunct but is not a definitive mortality-reducing therapy in this setting.