Sep 122020

Jon-Emile S. Kenny MD [@heart_lung]

As human beings, don’t we need questions without answers as well as questions with answers?”

-Alan Lightman

With the publication of REMAP-CAP, CAPE-COVID and CoDEX, the World Health Organization [WHO] Rapid Evidence Appraisal for COVID-19 Therapies [REACT] Working Group combined these results with 4 other randomized trials of corticosteroids in COVID-19.  The stated objective was to ‘estimate the association between administration of corticosteroids, compared with usual care or placebo, and 28-day all-cause mortality in hospitalized, critically ill patients with suspected or confirmed COVID-19.’

Accordingly, because the RECOVERY trial recruited both critically ill and non–critically ill hospitalized patients and it was not possible to distinguish whether patients had been critically ill but not receiving invasive mechanical ventilation at the time of randomization, only patients who received invasive mechanical ventilation were included in the analysis.

What They Did

Trials were located and assessed via a systematic search of, the Chinese Clinical Trial Registry, and the EU Clinical Trials Register. All recruiting clinical trials related to COVID-19 that examined the therapeutic efficacy of corticosteroids were identified.

Trials were categorized based upon which corticosteroid was employed and whether it was administered as a ‘low dose’ or a ‘high dose’ defined by the following cut-offs: 15 mg/day of dexamethasone, 400 mg/day of hydrocortisone, and 1 mg/kg/day of methylprednisolone.

The primary outcome that the REACT group identified was all-cause mortality up to 30 days after randomization.  Importantly, this was established before any outcome data were available from any of the studies.  As 5 trials reported mortality at 28 days after randomization, the primary outcome was reported as 28-day all-cause mortality; the secondary outcome was serious adverse events.

What They Found

Seven trials were included in the final meta-analysis.  The corticosteroid groups included dexamethasone at low and high doses, low-dose hydrocortisone, and high-dose methylprednisolone.  Three of the trials were very briefly outlined in part 1 while RECOVERY was considered previously.  In addition, the authors also included:

  1. The Efficacy of Dexamethasone Treatment for Patients with ARDS Caused by COVID-19 [DEXA-COVID 19, n = 19] which like CoDEX only enrolled mechanically-ventilated patients with moderate-to-severe ARDS.
  2. Steroids-SARI [n = 47] which included only patients in the ICU
  3. Hydrocortisone for COVID-19 and Severe Hypoxia [COVID STEROID, n = 29] which included patients receiving a minimum of 10 L/min of supplemental oxygen.

A total of 1703 patients were randomized [i.e. 1025 to usual care or placebo and 678 to corticosteroids] in the 7 trials.  As 1007 [59.1%] of the patients in the meta-analysis came from RECOVERY, the disproportionate randomization was entirely driven by the design of that trial [recall that patients were assigned to usual care or low-dose dexamethasone in a ratio of 2:1].

The summary odds ratio [OR] was 0.66 [95% CI, 0.53-0.82; P < 0.001] for all-cause mortality comparing corticosteroids with usual care or placebo.  This corresponds to an absolute mortality reduction of 8% [i.e. 32% with corticosteroids, assuming a mortality risk of 40% with usual care or placebo] or a number needed to treat [NNT] of 12.5.  This effect was largely propelled by dexamethasone given the large fraction of RECOVERY patients.  When considering only those who received hydrocortisone the OR was 0.69 [95% CI, 0.43-1.12; P = 0.13].

In patients who were invasively ventilated at randomization, the OR was 0.69 [95% CI, 0.55-0.86] while the OR was 0.41 [95% CI, 0.19-0.88] among patients who were not invasively ventilated, both favouring corticosteroids.  Of note, the OR was 0.86 [95% CI, 0.73-1.00] among those in the RECOVERY trial who required oxygen [with or without non-invasive ventilation] but were not invasively ventilated.

While the RECOVERY trial did not report adverse events, among the other trials, there was no difference between those randomized to corticosteroids or usual care/placebo.


The REACT meta-analysis followed my expectations delineated in part 1, though 59% of the meta-analysis was formed from RECOVERY, the trial that shaped my pre-conceptions.  In totality, the 7 trials support the use of at least low-dose corticosteroids in critically-ill patients, whether intubated or not with COVID-19-associated respiratory failure.  While the meta-analysis did not specifically investigate the progression of critically-ill, non-intubated patients to invasive ventilation, the OR for those without an endotracheal tube at randomization also showed reduced odds of death at 28 days.

Though not specifically addressed in part 1, another expectation of mine, gleaned from RECOVERY, was the hypothesis that symptom duration mediates corticosteroid efficacy.  That is, the benefit of anti-inflammation is maximized if administered more than a week after symptoms begin.  In SARS-CoV2, viral replication is generally greatest early in disease, while lung inflammation lags.  Thus, corticosteroids given ‘later’ are less likely to prolong viral replication and more likely to cool pulmonary inflammation and assuage abnormal gas exchange.  Indeed, this was found in REACT, which obviously echoes the RECOVERY data.  The OR was 0.64 [95% CI, 0.49-0.83] for patients who were symptomatic for more than 7 days prior to randomization and the OR was 0.63 [95% CI, 0.39-1.04] among those who were symptomatic for 7 days or less prior to randomization.  While hardly conclusive, it is certainly interesting.

Another mild surprise in the meta-analysis was the finding that corticosteroids had no clear mortality benefit in those on vasoactive medications as compared to those who were not.  Though not an objective of the REACT investigators, one wonders if the corticosteroids improved shock resolution and/or days free from cardiovascular support?  Nevertheless, this finding seemingly mirrors the ‘shock-dependent’ arm of REMAP-CAP.  It suggests, that the beneficial mechanism of corticosteroids in critically-ill COVID-19 patients is not primarily hemodynamic in origin.

Lastly, it is reassuring that there was no clear harm signal in the trials reporting adverse events.  Thus, speculation that mineralocorticoid activity engendered by hydrocortisone could be uniquely harmful in SARS-CoV2 infection is largely extinguished.

Yet as the thoughtful, accompanying editorial highlights, many questions remain.  One in particular which also clangs around my cranium is the question of ARDS sub-types?  Might hyper-inflammatory ARDS benefit more?  If so, do we need to also consider symptom duration so as to exclude steroids when viral replication is high?  What about patients with ‘L-type CARDS’ who have high oxygen requirements but do not meet the Berlin criteria for moderate-to-severe ARDS?



Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he also the creator and author of a free hemodynamic curriculum at heart-lung.orgPlease download my free textbook.


Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


More on Corticosteroids in COVID-19 – Part 2: REACT meta-analysis