October 2020, Volume XXXIV, Number 6

  Infectious Disease

Treating COVID-19 with corticosteroids

Positive worldwide collaboration

espiratory illness caused by the novel coronavirus SARS-CoV-2 (now known as COVID-19) was first identified in the Chinese city of Wuhan and surrounding Hubei Province in early December 2019. The first U.S. and Minnesota cases were identified on January 20 and March 6, respectively. On March 11 the World Health Organization (WHO) characterized COVID-19 as a global pandemic.

From the outset, Chinese clinicians who were inundated with large numbers of seriously ill patients attempted a variety of therapies, including antibiotics, antivirals, antioxidants, N acetyl-L-cystine, and hydroxychloroquine. In addition, some of the most severely ill patients were given corticosteroids. An early report from Wuhan suggested that corticosteroids might decrease mortality of patients with COVID-related ARDS (acute respiratory distress syndrome), but interpretation of this finding was limited by the retrospective, non-randomized study design.

We began to use corticosteroids routinely in nearly all mechanically ventilated patients

Indeed, recommendations from the WHO, Centers for Disease Control and Prevention (CDC), German Intensive Care Society, American Thoracic Society, and the Infectious Diseases Society of North America recommended against steroid administration in patients with COVID-19 infection outside of a randomized clinical control trial. These recommendations were based on previous studies that had raised concerns about potential harm related to use of corticosteroids in the management of other respiratory viruses, including influenza and the coronaviruses responsible for SARS and MERS. Corticosteroid therapy appeared to decrease viral clearance in SARS and MERS and possibly increase mortality of influenza. In contrast, the Chinese Thoracic Society recommended steroids be given to critically ill patients with COVID-19 and the Society of Critical Care Medicine made a weak recommendation in favor of corticosteroids for COVID-19-related ARDS. These conflicting recommendations regarding use of corticosteroids created a dilemma for front-line clinicians responsible for managing patients with COVID-19. In the absence of evidence from randomized controlled trials, individual physicians were forced to use their own clinical judgement regarding the risk-benefit of treatment with steroids.


The controversy regarding use of corticosteroids in patients with ARDS has been ongoing for decades. In a large multicenter trial, steroids given to patients with late (after day 5) ARDS did not affect outcome, and one meta-analysis of previous studies found no overall benefit to the use of steroids. Previous investigations varied widely with the specific type of steroid, dose, and duration of therapy. Earlier studies were also conducted without the application of lung-protective ventilation, an intervention associated with improved outcomes in ARDS. A subsequent meta-analysis that was limited to studies with low-to-moderate steroid dosing, a duration of 1–4 weeks of administration, and lung protective ventilation concluded that steroid administration in ARDS increased ventilator-free days and decreased mortality. Recently, a multicenter randomized trial (DEXA ARDS) that assessed the impact of a 10-day course of dexamethasone (20 mg/day for 5 days, then 10 mg/day for 5 days) when given within 48 hours of onset of ARDS, also found that dexamethasone was associated with an increase in ventilator free-days and decreased 60-day mortality (Villar, March 2020 Lancet Respiratory Medicine).

Our approach

At Hennepin Healthcare, rather than take an all-or-nothing approach, we made the initial decision to give corticosteroids only to COVID-19 patients with severe ARDS. Over time, as we observed what appeared to be a favorable response in some cases, we began to use corticosteroids routinely in nearly all mechanically ventilated patients with ARDS and for those patients on the ward who were requiring at least 6 liters/minute of oxygen. The latter approach was prompted by an observational study from Henry Ford Health System that suggested use of steroids earlier in the course of COVID-related respiratory illness might decrease the risk of progression to overt respiratory failure. At this point in time, however, reliable data regarding the risk vs. benefit of steroids in COVID 19 was lacking and it was uncertain if this approach was justified.

Fortunately, researchers in the United Kingdom were able to rapidly organize and complete a large randomized trial of the use of steroids for COVID-19. Preliminary results were published in mid-June, only a few months after the pandemic had reached the United States and Europe (RECOVERY Collaborative Group, July 2020 New England Journal of Medicine). This impressive study recruited 11,500 patients from 175 NHS hospitals who were randomized to various treatment arms, with preliminary results announced 98 days after protocols were drafted. A total of 2,104 patients hospitalized with COVID-19 were randomized to receive 10 days of dexamethasone (6 mg/day) for 10 days. As compared to placebo, dexamethasone resulted in a highly significant reduction in mortality (29% vs. 41%) for those patients who required mechanical ventilation. For non-intubated patients who required supplemental oxygen, the reduction in mortality (23% vs. 26 %) was less pronounced but still statistically significant.

Corticosteroids have been shown to significantly reduce the mortality of critically ill patients

In contrast, among patients who did not require supplemental oxygen, dexamethasone did not provide benefit and there was a trend toward worse outcomes. With publication of this landmark study, other placebo-controlled trials of steroids for critically ill patients with COVID-19 suspended enrollment, because it was deemed unethical to withhold steroids. A recent meta-analysis by a WHO committee of experts examined available data from six of the latter studies and the subset of RECOVERY patients who had required mechanical ventilation (Sterne, September 2020 JAMA). They concluded there was unequivocal evidence in favor of steroids for critically ill patients with COVID-19, but did not make a recommendation as to the type of steroid, dose, or duration of treatment that should be used. Importantly, they found that corticosteroids did not increase risk of infections or other serious adverse events. For mechanically ventilated patients with COVID-related ARDS, our current practice is to give 20 mg dexamethasone a day for 5 days followed by 10 mg a day for another 5 days. For non-intubated patients who require at least 2 liters of supplemental oxygen, we use a somewhat lower dose of 10 mg a day for 5 days followed by 6 mg a day for 5 days. COVID-19 patients who do not require oxygen are not given corticosteroids.

Commenting on the results

The fact that corticosteroids appear to have their greatest benefit for the most severely ill COVID-19 patients, but not for those with milder illness, is consistent with what is known about the evolution of COVID-19 respiratory illness. The clinical course of COVID can be divided into two phases: the viral replicative phase and a subsequent hyper-inflammatory immune response phase (see Figure 1).

Figure 1. COVID-19 disease phases and indicated therapy.

Many patients with COVID-19 remain asymptomatic or have very mild illness and presumably never progress beyond the viral replicative phase. When there is progressive and clinically significant respiratory involvement, patients often present to the hospital within 5–10 days after the onset of symptoms. Respiratory failure may be present at the time of admission, or may develop after hospitalization. Critically ill patients with COVID-19 usually have evidence of hypercoagulability and elevated inflammatory markers believed to be related to various proinflammatory cytokines. Therapies that target viral replication, such as remdesivir—and possibly convalescent plasma—are of most benefit when given early in the clinical course of COVID, but are likely of little value when patients have developed a hyper-immune inflammatory response with severe respiratory failure. In contrast, corticosteroids appear to be of greatest benefit when excessive inflammation has resulted in respiratory failure, but not earlier in the disease when viral replication is dominant.


Corticosteroids have been shown to significantly reduce the mortality of critically ill patients with COVID-19-related ARDS who require mechanical ventilation. For non-intubated patients with COVID 19, corticosteroids should also be given to those who need supplemental oxygen but not to those who have adequate oxygen saturation.

James W. Leatherman, MD,  is a member of the Division of Pulmonary and Critical Medicine at Hennepin County Medical Center.

John B. Pflugi, DO,  is a current critical care fellow at Hennepin County Medical Center. 


PO Box 6674, Minneapolis, MN 55406

(612) 728-8600


follow us

© Minnesota Physician Publishing · All Rights Reserved. 2019


John B. Pflugi, DO,  is a current critical care fellow at Hennepin County Medical Center.