Table 2

GRADE assessment

Certainty assessmentNumber of patientsEffectCertaintyImportance
Number of studiesStudy designRisk of biasInconsistencyIndirectnessImprecisionOther considerationsSofosbuvir/daclatasvirControlRelative
(95% CI)
Absolute
(95% CI)
Mortality
3Randomised trialsSerious*Not seriousNot seriousNot seriousPublication bias strongly suspected strong association†5/92 (5.4%)17/84 (20.2%)RR: 0.31
(0.12 to 0.78)
140 fewer per 1000
(from 178 fewer to 45 fewer)
⨁⨁⨁◯
Moderate
High
Need for ICU/IMV
3Randomised trialsSerious*Not seriousNot seriousNot seriousPublication bias strongly suspected strong association†9/92 (9.8%)24/84 (28.6%)RR: 0.35
(0.18 to 0.69)
186 fewer per 1000
(from 234 fewer to 89 fewer)
⨁⨁⨁◯
Moderate
High
Clinical recovery
4Randomised trialsSerious*Not seriousNot seriousNot seriousPublication bias strongly suspected‡105/118 (89.0%)82/112 (73.2%)RR: 1.20
(1.04 to 1.38)
146 more per 1000
(from 29 more to 278 more)
⨁⨁◯◯
Low
High
  • *One study is not a randomised trial; two studies are open label.

  • †All studies originated from a single country. Small-sample size with all positive studies.

  • ‡All studies originated from a single country. Small-sample size with all positive studies except for one.

  • GRADE, Grading of Recommendations Assessment, Development and Evaluation; ICU, intensive care unit; IMV, invasive mechanical ventilation; RR, Risk ratio.