Randomised Controlled Trial (RCT)

A randomised controlled trial is a prospective experiment in which participants are allocated by chance to two or more arms, at least one of which serves as control. Because allocation is independent of prognosis, the arms are comparable at baseline on measured and unmeasured factors alike, so a difference in outcome can be attributed to the intervention rather than to who received it. What it estimates is an average treatment effect in the population studied, not a prediction for any individual.

The label alone guarantees little, because the strength of an RCT lies in its execution: concealed allocation, blinding wherever feasible, prospective registration, analysis by intention to treat, and follow-up complete enough that missing data cannot drive the result. The obesity and diabetes incretin programmes are large well-run examples of the form. By contrast, most claims circulating for research peptides have no randomised trial behind them at all, which is a statement about the evidence rather than about the molecules.

An RCT answers whether this intervention changed this outcome in this population, which is internal validity. It does not automatically answer what happens in a clinic. Restrictive eligibility criteria, run-in periods, protocol-driven adherence and intensive monitoring all make trial participants unlike the people later prescribed the drug, so the effect seen in practice is usually smaller.

Treating RCT as a synonym for reliable is the standing error. A small unregistered open-label randomised study with a subjective endpoint and heavy dropout can easily be weaker evidence than a large well-adjusted cohort study. In the other direction, a single positive trial is one result and not a literature; replication and systematic review exist because individual trials, including good ones, disagree.