Find the Strongest Evidence

What the best evidence shows

ACTT-1, a randomized double-blind placebo-controlled trial, found a 4-day shorter median recovery time with remdesivir: 11 days versus 15 days among 1,063 hospitalized adults. Its mortality estimates favored remdesivir, but the clearest and most reproducible outcome was faster recovery rather than definite survival improvement. ACTT-1 efficacy and design

Results vary by disease stage. The 2025 CADTH synthesis reports that the strongest randomized-trial signal was among patients receiving low-flow oxygen; benefit was not consistently demonstrated in patients with no oxygen requirement or advanced mechanical ventilation. In WHO Solidarity, overall 28-day in-hospital mortality was not clearly reduced (RR 0.91, 95% CI 0.82–1.02), whereas patients on oxygen but not ventilated showed a possible reduction (RR 0.87, 95% CI 0.76–0.99). CADTH evidence synthesis

How large is the apparent effect?

A 2025 meta-analysis combining 21 randomized trials with 101 real-world studies estimated lower mortality overall (OR 0.69, 95% CI 0.55–0.86) and approximately 1.84 fewer days to clinical improvement. However, only four randomized trials contributed to the overall mortality meta-analysis, while most participants came from observational studies; therefore, the pooled mortality estimate is more vulnerable to confounding and sponsor-related influence than the randomized recovery result. The same analysis found no clear overall difference in time to clinical recovery (MD −0.31 days, 95% CI −3.37 to 3.94). 2025 systematic review and meta-analysis

Interpretation and blind spots

• Most defensible conclusion: remdesivir can shorten recovery for some hospitalized patients, with the clearest signal around low-flow oxygen; a universal mortality benefit is not established. Population-specific interpretation
• Evidence from early-pandemic trials may not transfer exactly to people with later variants, different prior immunity, or contemporary hospital care. Observational studies can also be distorted by treatment-selection, timing, residual-confounding, and immortal-time biases. Evidence limitations and conflicts

Confidence: high that the drug can accelerate recovery in a subset of hospitalized patients; moderate-to-low that it reduces mortality broadly, because randomized mortality findings are inconsistent and pooled estimates rely heavily on observational evidence.