No expert claims captured yet — this verdict will fill in as we index more sources.
Mechanistic and trial evidence converge on a real, replicable effect.
Mechanistic and trial evidence converge on a real, replicable effect.
Benefits hold across the populations where it's been tested.
Benefits hold across the populations where it's been tested.
The effect size is large enough to matter clinically, not just statistically.
Confounding and publication bias inflate the apparent benefit.
The headline effect shrinks once you account for trial quality.
The headline effect shrinks once you account for trial quality.
Animal-model results don't translate to the human protocol being recommended.
The headline effect shrinks once you account for trial quality.
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.