13 Research methods, statistics, critical review and evidence-based practice13.3 Experimental studies
13.3.8 Clinical trial phases
2 min read
Written by Resident Doctor Committee · Published
New in the 2027 syllabus
The College added this topic for exams from January 2027.
The four main phases
Each phase asks a different question.
Phase | Main question | Typical focus |
|---|---|---|
I | What doses can be given, and how does the body handle them? | Safety, tolerability, pharmacokinetics and dose escalation (stepwise dose increases) |
II | Are there early signs that it works in the intended condition? | Choosing the dose, early evidence of efficacy and more safety data |
III | In larger studies, do the benefits outweigh the harms? | Confirmatory comparisons that support licensing decisions |
IV | What happens after marketing authorisation? | Effectiveness in wider use, uncommon harms and longer-term use |
Phase I may recruit healthy volunteers or people with the condition. The choice depends on the intervention and its risks.
Phase II results can justify a larger trial. They usually cannot show uncommon harms, or how well the treatment works across routine services.
Phase III trials usually randomise participants against placebo or an active comparator (an existing treatment). The design still decides which conclusions the results support.
Phase IV covers research after authorisation, including both trials and observational studies.
Where the labels blur
Phase 0 studies are exploratory. They use limited early exposure to answer biological or pharmacokinetic questions, not to test whether a treatment works.
Combined phases, such as phase I/II, join the aims of two phases. The label is no substitute for reading the protocol.
Study size depends on the condition and the purpose of the study. A fixed number of participants does not define a phase.
Marketing authorisation weighs the submitted evidence for specified uses. It does not prove that all harms are known, or that the treatment is cost-effective.
Appraising the evidence
Limit any claim about safety to the participants, doses and length of observation actually studied.
Rare or delayed harms may only appear after licensing, because early studies expose too few people for too short a time.
Post-marketing surveillance can detect warning signals. Spontaneous reports on their own cannot estimate how often a harm occurs, because they lack the complete exposure denominator (the total number of people who have taken the treatment).