Table 2.
GRADE summary of findings: corticosteroids in patients with COVID-19, based on direct evidence from observational studies and one randomized controlled trial of patients with COVID-19.
| Outcomes | Population | Relative effects (95%CI) | Baseline risk for control group | Absolute risks difference (95% CI) | Quality of evidence | Summary |
|---|---|---|---|---|---|---|
| Mortality | COVID-19 with different disease severity | RR 0.92 (0.69–1.22) based on data from 10,278 patients in 9 observational studies and 1 RCT | 19.9% | −1.6% (−6.2% to 4.4%) | Very low (serious inconsistency and imprecision) | We are very uncertain of the effect of corticosteroids on short term mortality in patients with COVID-19 |
| Mortality | Critical COVID-19 | RR 0.80 (0.26–2.46) based on data from 1719 patients in 2 observational studies and 1 RCT | 52.2% | −10.4% (−38.6% to 76.2%) | Very low (serious inconsistency) | We are very uncertain of the effect of corticosteroids on short term mortality in patients with critical COVID-19 |
| Mortality | Severe COVID-19 | RR 0.98 (0.73–1.30) from 9673 patients in 6 observational studies and 1 RCT | 15.8% | −0.3% (−4.3% to 4.8%) | Very low (serious inconsistency) | We are very uncertain of the effect of corticosteroids on short term mortality in patients with severe COVID-19 |
| Mortality | Non severe COVID-19 | RR 0.67 (0.19–2.34) based on data from 1824 patients in 1 observational study and 1 RCT | 14% | −4.6% (−11.4% to 18.8%) | Very low (serious inconsistency) | We are very uncertain of the effect of corticosteroids on short term mortality in patients with non-severe COVID-19 |
| Composite outcome of death/ICU/MV | Severe COVID-19 | RR 0.41 (0.23–0.73) based on data from 676 patients in 4 observational studies | 44.1% | −26% (−33.9% to 11.9%) | Very low (serious inconsistency) | We are very uncertain of the effect of corticosteroids on composite outcome of death/ICU admission/MV in patients with severe COVID-19 |
| Mechanical ventilation need | Severe COVID-19 | RR 0.74 (0.50–1.09) based on data from 5768 patients in 2 observational studies and 1 RCT | 9.3% | −2.4% (−4.7% to 0.8%) | Very low (serious inconsistency) | We are very uncertain of the effect of corticosteroids on mechanical ventilation need in patients with severe COVID-19 |
| Delayed viral clearance | COVID-19 with different disease severity | RR 1.47 (1.11–1.93) based on data from 760 patients in 6 observational studies | 20.7 % | 9.7% (2.3%–19.2%) | Low | Corticosteroids may delay viral clearance in patients with COVID-19 |
| Acquired blood stream infections | COVID-19 with different disease severity | HR 3.95 (1.20–13.03) based on data from 78 patients in 1 observational study | 45.5% | 45.4% (6.2%–54.5%) | Very low (serious inconsistency) | Corticosteroids probably do increase the risk of acquired blood stream infections in patients with COVID-19 |