SCCM Guideline Update: Steroids for Community-Acquired Pneumonia, ARDS, Septic Shock
SCCM updates its recommendations for steroids in critically ill patients, alone
Steroids are good medicine in the ICU, U.S. critical care professional societies agree.
The Society of Critical Care Medicine (SCCM) expanded its prior 2017 guidance, now recommending strongly that corticosteroids be given for severe community-acquired pneumonia, and also suggesting steroids for patients with acute respiratory distress syndrome (ARDS) and septic shock of any severity. The new advisement comes in the wake of randomized trials published since 2017.
SCCM made its statement unilaterally, without the European Society of Intensive Care Medicine (ESICM), its partner on the 2008 and 2017 updates, or the American Thoracic Society (ATS).
Background
SCCM and ESICM together coined the phrase critical illness-related corticosteroid insufficiency (CIRCI) back in 2008—and it promptly did not catch on.
That could be because CIRCI is ill-defined, almost hypothetical. There has never been an accepted or reliable way to diagnose the condition. Circulating cortisol and response to cosyntropin are not useful as routine tests: many “steroid-responsive” patients have normal or elevated serum cortisol, and/or have negative cosyntropin stimulation tests.
The purportedly inadequate glucocorticoid effects occur intracellularly, where they can’t be measured today. CIRCI manifests, it is said, in the elevated and dysregulated immune/inflammatory activity (elevated cytokines, coagulation tests, etc.) observed during critical illness.
In the absence of accepted diagnostic criteria, practically speaking, the diagnosis of “corticosteroid insufficiency” is made at the moment an attending says “I think we should give this patient steroids.”
The hypothalamic-pituitary-adrenal axis and cortisol are intimately involved in critical illness and recovery, but how remains poorly understood. Critically ill patients with low cortisol (e.g., total cortisol levels < 10 μg/dL) experience higher mortality, but that may be more a marker of severity of illness than a remediable deficiency. In mechanically ventilated patients with septic shock, randomized trials have not shown improved survival with steroid administration, even among those who were “cosyntropin responders” (although hemodynamics were improved with steroids).
Recent studies, though, have shown steroids’ benefit in Covid pneumonia with ARDS, and for steroids improving outcomes in severe community-acquired pneumonia.
The new evidence prompted all three societies to issue new guideline updates. Breaking with their past pattern of joint statements, ATS, SCCM, and ESICM all acted separately this time.
What follows are PulmCCM’s interpretations of the SCCM panel’s statement, which you can read in its entirety here. PulmCCM is not affiliated with SCCM, ATS, or ESICM.
Give Steroids for Severe Community-Acquired Pneumonia
This was a strong recommendation based on convincing evidence, strengthened from the panel’s 2017 suggestion to provide steroids for hospitalized patients with community-acquired pneumonia.
Severity of pneumonia was not formally defined in the recommendation, owing to the heterogeneity of analyzed trials, but mechanical ventilation, severe sepsis, high pneumonia severity scores, ICU admission, or need for vasopressors were some heuristics used.
Drug choice and dosing were left up to the clinician, with methylprednisolone 40 to 80 mg/d IV for 5–7 days (or equivalent steroid agents/doses) mentioned as a guidepost. Examples from clinical trials included longer treatment courses and tapers:
Methylprednisolone 40 mg IV loading dose, then 40 mg/day for 7 days, 20 mg/day for 7 days, 12 mg/day for 3 days, 4 mg/day for 3 days
Hydrocortisone 200 mg IV daily (continuous infusion) for up to 8 days, according to clinical improvement, then tapered for 8 days.
Consider Steroids for All Patients With ARDS
Like ATS, SCCM now suggests considering giving corticosteroids to all patients with ARDS, regardless of severity. It’s an expansion of the 2017 guidance in which SCCM/ESICM suggested steroids for moderate or severe ARDS.
The evidence supporting the change rests primarily on the apparent benefit of steroids noted in ARDS due to Covid-19 (over half of the patients in the included trials) and community-acquired pneumonia. Prior studies (pre-Covid, enrolling patients with ARDS of all causes) were inconclusive and conflicting, but meta-analyses suggested a potential benefit from steroids.
A large proportion of ICU patients have abnormal chest films and hypoxemia. Because the threshold for diagnosis of mild ARDS (PaO2/FiO2 ratio of <300 + bilateral infiltrates not due to heart failure) is so inclusive, closely following this new suggestion would result in steroid treatment in an enormous new cohort of patients.
No drug or dosing regimen was advised over another. Dexamethasone 20 mg IV daily for 5 days, followed by 10 mg IV daily for 5 days or until extubation, was one option mentioned. They cited weak evidence suggesting treatment courses longer than 7 days may be associated with lower mortality (compared to <7 days).
The professional societies are going solo on recent guideline updates, especially for ARDS. In 2017, ATS, SCCM, and ESICM together issued a joint guideline for the management of ARDS. Then in 2023, ATS and ESICM released separate, conflicting ARDS guidelines: ESICM did not recommend steroids for ARDS in its guidance, while ATS did. SCCM joined neither ATS nor ESICM’s 2023 ARDS guideline update. Yes: it’s complicated and confusing. We covered those details and more here:
Steroids Now Suggested for All Septic Shock (Not Just “Refractory”)
The SCCM panel expanded its prior guidance, now suggesting giving moderate-dose intravenous corticosteroids for all patients with septic shock. The 2017 SCCM/ESICM suggestion had been to use steroids for severe septic shock (requiring high-dose or multiple vasopressors).
Since that 2017 update, two large trials were published, with one (lead author: chair of the SCCM panel) showing a mortality benefit—possibly the first ever in a trial this size—and the much larger ADRENAL (by ANZICS) showing no mortality benefit, but faster resolution of shock with steroids.
Hydrocortisone 200-300 mg IV/day in divided doses for five to seven days (or until ICU discharge) was an often-used treatment course in analyzed trials. There was insufficient evidence to advise for or against co-administration of mineralocorticoids like fludrocortisone (a recent trial suggested this could be helpful). Co-administration of mineralocorticoid might include fludrocortisone 50 µg enteral daily.
The evidence that steroids improve mortality in septic shock is low, and if present, any such effect is small. In multiple large randomized trials testing steroids in septic shock, no survival benefit was noted with steroids. Combining 46 randomized trials analyzed, a relative risk of 0.94 for mortality with steroids was gleaned, with a confidence interval including 1.0. Considering the possibility for amplification of publication bias or other biases in the aggregate analysis, this is inconclusive (hence the weak recommendation).
On the other hand, septic shock patients receiving steroids in randomized trials consistently tend to require shorter durations of vasopressors and have lower organ failure scores, and seem to better survive their ICU stays. The authors expressed hope that by expanding steroid use globally for septic shock, potential small benefits would aggregate into large absolute benefits, without excess harm from steroid side effects. They acknowledge steroids’ risk of long-term neuromuscular weakness has not been well-studied, but this risk was considered acceptable when using moderate doses for short periods.
High-dose steroids were advised against in septic shock due to potential harm, even when limited to shorter periods (e.g., 3 days).
Conclusion
In its new guideline update, SCCM now strongly recommends moderate-dose corticosteroids be given for patients with severe community-acquired pneumonia. Previously, this was a weak recommendation.
In addition, steroids are now suggested for all patients with ARDS or septic shock. In the 2017 guidance, steroids were only suggested for patients with moderate-to-severe ARDS, and for septic shock requiring high-dose or multiple vasopressors.
The past consensus between the major professional societies seems to have fractured over the issue of whether and when to give steroids during critical illness. Instead of the usual practice of making such guideline statements together, SCCM did so here without its prior co-authors in ESICM.
For its part, ATS also recently suggested steroids be considered for all ARDS patients, but did so without its previous cosignatories in SCCM or ESICM—the latter issuing its own ARDS guideline update, without mentioning steroids at all.
The fragmented guidance reduces its perceived authoritativeness, and the weakness of the evidence (and the recommendations themselves) for steroids in ARDS and septic shock together keep the decision-making in the hands of the clinician.
Or as an influential intensivist posted after ATS-SCCM-ESICM’s confusing guidance schism, the “silver lining” is “I can now justify giving steroid to almost anyone because they have ‘ARDS’.”
A little bit of steroids, white wine and morphine doesn't hurt anyone...