Three phases, three different questions | Visual AI Learning Guide

The three phases are not three attempts at the same question. They are three different questions asked in order. Phase I is about safety, in a small group, often healthy volunteers. Phase II is the first real test of whether the drug helps patients, in a few hundred people with the condition. Phase III is the confirmation step: thousands of patients, usually compared against the current standard treatment, to be sure the benefit is real and to catch harms too rare to appear earlier. Read the panels left to right and watch the size grow while the question gets stricter. Each panel is a gate, and a drug that fails one never reaches the next.

0:00 / 0:00

Auto-play

Once a candidate enters human testing, the trials are split into three phases, and each phase answers a different question. Phase I asks whether the drug is safe. Phase II asks whether it appears to work. Phase III asks whether it works better than what is already used, in a large enough group to be confident.

The phases also differ in size. Phase I involves a small number of participants, often healthy volunteers, and focuses on safety and how the body handles the drug. Phase II involves a few hundred patients with the condition and looks for an early signal of benefit. Phase III involves thousands of patients and usually compares the new drug against the current standard treatment, both to confirm benefit and to detect harms that are too rare to show up in smaller groups.

Each phase is a filter. A drug that fails Phase I never reaches Phase II, and a drug that fails Phase III has already consumed years of work and large sums of money. That is why the later a failure occurs, the more damaging it is.