Skip to main content
. 2022 Jan 24;48(1):e20210393. doi: 10.36416/1806-3756/e20210393

Table 3. GRADE analysis of remdesivir compared with placebo/standard of care in patients hospitalized with moderate to severe COVID-19 in randomized clinical trials.

Certainty assessment No. of patients Effect Certainty Importance
No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations Remdesivir Placebo/SOC Relative (95% CI) Absolute (95% CI)
Mortality in 29 days
5 Randomized trials very seriousa not serious not serious not serious none 390/3669 (10.6%) 397/3543 (11.2%) RR 0.94 (0.82 to 1.07) 7 fewer per 1,000 (from 20 fewer to 8 more) MODERATE IMPORTANT
Recovered patients in 29 days
3 Randomized trials very seriousb not serious not serious very seriousc none 579/768 (75.4%) 525/757 (69.4%) RR 1.09 (1.03 to 1.15) 62 more per 1,000 (from 21 more to 104 more) LOW IMPORTANT
Clinical improvement in 29 days
2 Randomized trials very seriousd not serious not serious very seriousc none 277/351 (78.9%) 211/278 (75.9%) RR 1.10 (1.01 to 1.19) 76 more per 1,000 (from 8 more to 144 more) LOW IMPORTANT
Mechanical ventilation or ECMO in 29 days
5 Randomized trials very seriousa very seriouse not serious not serious very seriousf 354/3268 (10.8%) 375/3152 (11.9%) RR 0.76 (0.46 to 1.26) 29 fewer per 1,000 (from 64 fewer to 31 more) VERY LOW IMPORTANT
very serious adverse events in 29 days
3 Randomized trials not serious not serious not serious very seriousc None 169/880 (19.2%) 201/794 (25.3%) RR 0.75 (0.63 to 0.90) 63 fewer per 1,000 (from 94 fewer to 25 fewer) MODERATE IMPORTANT

GRADE: Grading of Recommendations Assessment, Development and Evaluation; SOC: standard of care; and RR: risk ratio.

a. Three of the five studies were not blinded by the researcher. Two studies showed risk of bias related to randomization and allocation, with some concerns.

b. Two of the three studies were not investigator’s blinded studies, and one study had a risk of bias for randomization and blind allocation, with some concerns.

c. Large confidence interval

d. One of the two included studies was not blinded by the researcher, and one had a risk of bias for randomization and blind allocation with some concerns.

e. High heterogeneity (I2 = 71%).

f. Publication bias.