What is a phase-3 trial and why does it matter?

Clinical trial development follows a standard four-phase structure. Phase 1 tests safety and pharmacokinetics in a small number of healthy volunteers (typically 20–100). Phase 2 tests efficacy and dose-finding in a moderate patient population (typically 100–500). Phase 3 tests efficacy and safety at scale in a large patient population (typically 1000–5000) using randomised, controlled, blinded design. Phase 4 is post-market surveillance after approval.

The phase-3 trial is what regulatory agencies typically require for approval because its design controls for the largest number of confounders. Randomisation controls for baseline differences between arms. Placebo control controls for placebo effects and natural history. Double-blinding controls for observer bias and expectation effects. Large sample size gives statistical power to detect clinically meaningful differences and to observe rare adverse events. Pre-specified endpoints prevent post-hoc analysis inflation of false positives.

When a compound page cites 'STEP-1 (Wilding 2021 NEJM)' as pivotal evidence, that reference identifies a specific phase-3 trial: sample size in the thousands, defined primary endpoint (percent weight loss at 68 weeks), randomised placebo-controlled double-blind design, published in a top-tier peer-reviewed journal. That combination is the highest standard of evidence for demonstrating that a compound produces a specific clinical effect at scale. Compounds without this evidence tier can still be scientifically interesting, but claims about clinical effect require weaker evidence to be interpreted more cautiously.