Effect of 12 mg vs 6 mg of Dexamethasone on the Number of Days Alive Without Life Support in Adults With COVID-19 and Severe HypoxemiaThe COVID STEROID 2 Randomized Trial

Question – 

What is the effect of 12 mg vs 6 mg of dexamethasone on the number of days alive without life support at 28 days inpatients with COVID-19 and severe hypoxemia?

Setting and design

The COVID-STEROID-2 randomized controlled trial was conducted between August 27, 2020, and May 20, 2021, including 11 centres in Denmark, 12 in India, two in Sweden, and one in Switzerland. Stratified randomization by trial site, age less than 70 years, and requirement for invasive ventilation at screening. Patients were randomized in a 1:1 ratio to receive 6 vs. 12 mg dexamethasone.

Study population

Inclusion

Patients with confirmed SARS-CoV-2 infection who required oxygen at a flow rate of at least 10 L/min, NIV or CPAP, or were intubated and ventilated.

Exclusion

Glucocorticoids in doses higher than 6 mg of dexamethasone or equivalents for non-covid indications; on glucocorticoids for COVID-19 for 5 days or longer; the presence of invasive fungal infection or active tuberculosis; history of hypersensitivity reactions to dexamethasone; pregnancy

Interventions

Low-dose arm

Dexamethasone 6 mg IV up to 10 d after randomization

High-dose arm

Dexamethasone 12 mg IV up to 10 d after randomization

Other treatment modalities were left to the best judgement of clinicians. Other immunosuppressive agents were discouraged. Tocilizumab was allowed from January 2021 after the publication of the REMAP-CAP trial results.

Sample size

1000 patients for a 15% relative reduction in the 28-day mortality, 10% relative reduction in the time requiring life support, with 85% power.

1414 were screened, 414 were excluded

62% patients were recruited from Europe and 38% from India

Baseline characteristics

12 mg vs. 6 mg

Supplemental oxygen 55 vs. 53%

NIV or CPAP 24 vs. 26%

Invasive ventilation 22 vs. 20%

Other baseline characteristics were similar, except diabetes, higher in the 6 mg group.

Primary outcome

Composite outcome: alive and free of life support (free of invasive ventilation, circulatory support, and RRT)

Median number of days alive without life support at 28 days after randomization: 22.0 vs. (IQR, 6.0-28.0) 20.5 days (IQR, 4.0-28.0) days (adjusted mean difference, 1.3 days [95% CI, 0-2.6 days], P = 0.07

Alive and free of life support as percentage of patients: 42.6% vs. 40.2%

Individual components of the composite primary outcome

Days alive without invasive mechanical ventilation

Days alive without circulatory support

Days alive without RRT

No difference on adjusted analysis (adjusted for baseline comorbidities including ischemic heart disease or heart failure, diabetes, chronic obstructive pulmonary disease, use of immunosuppressive therapy within the prior 3 months, use of circulatory support, and use of kidney replacement therapy)

Secondary outcomes

No. of days alive without life support at 90 d, median: 84 vs. 80 days

No. of days alive and out of hospital at 90 d, median: 61.5 vs. 48 days

28-d mortality: 27.1% vs. 32.3% (not statistically significant)

90-d mortality: 32% vs. 37.7% (not statistically significant)

Serious adverse events (not different)

New episodes of septic shock

Invasive fungal infection

Clinically important gastrointestinal bleeding

Anaphylactic reaction to dexamethasone

Strengths – 

1. Pragmatic trial
2. A relatively large number of patients
3. Conducted across 2 different continents – good external validity
4. Blinding, allocation concealment and good follow up rate – good internal validity

Limitations –

• The presumed difference in outcomes between groups (15% relative reduction in the 28-day mortality, 10% relative reduction in the time requiring life support) may have been too large; a trend towards improved 28- and 90-d survival. The sample size may have been too small; hence the possibility of type II error.
• Composite primary outcome – underpowered
• The frequentist approach itself may be inferior to the bayesian approach for such a trial.
• Mortality among subgroups: especially at different levels of respiratory support may have been useful
• Concomitant use of tocilizumab halfway through may have confounded the results.

Take- home – 

This trial has added significant information on the role of steroids in COVID-19 however, it fails to change the current practices.