A clinical trial is a prospective study that evaluates the effects of a medical intervention — a medicine, device, procedure or behavioural change — in human participants under a pre-specified protocol. Trials are organised into phases, each answering a different question and building on the evidence of the last. This article describes the structure and governance of trials from a methodology and standards perspective; it is not clinical advice.
The four phases
| Phase | Primary question | Typical scope |
|---|---|---|
| Phase I | Is it safe, and how is it handled by the body? | First-in-human; small numbers; focus on safety, tolerability and dose-finding. |
| Phase II | Does it show signs of working, and at what dose? | Larger groups; preliminary efficacy and further safety. |
| Phase III | Does it work better than current options? | Large, often multi-centre randomised controlled trials supporting regulatory approval. |
| Phase IV | How does it perform in routine use? | Post-marketing surveillance after approval; rare effects and long-term outcomes. |
How trials are designed
The strongest designs use randomisation to allocate participants to groups, blinding to reduce expectation bias, and a control group — often a placebo or an existing standard of care — for comparison. A pre-registered protocol specifies the hypotheses, primary and secondary outcomes, sample size and analysis plan before data are collected, which guards against selective reporting. These ideas connect directly to our explainers on the placebo and placebo effect and randomised controlled trials.
Registration and transparency
Clinical trials are expected to be registered in a public registry — such as ClinicalTrials.gov or an ISRCTN registry — before they begin. The International Committee of Medical Journal Editors (ICMJE) requires prospective registration as a condition of publication, and the World Health Organization maintains a registry network and a minimum data set. Registration creates a public record of what a trial set out to do, so that its results can be checked against its original aims and so that unpublished trials do not vanish from the evidence base.
Governance and ethics
Trials are governed by independent research ethics committees (institutional review boards), by informed consent from participants, and by adherence to Good Clinical Practice. International principles trace back to the Declaration of Helsinki. Data are monitored, adverse events reported, and the conduct of the trial audited. Reporting is governed by the CONSORT statement for randomised trials, which specifies what must appear in the published account.
Why this matters to the research record
A registered protocol, a transparent results report and a persistent identifier together make a trial part of a citable, auditable record. The same contributor-attribution and identifier infrastructure that CASRAI works on — ORCID for people, DOIs for outputs, registries for studies — is what lets the scholarly and regulatory records stay connected.
Frequently asked questions
What are the four phases of a clinical trial?
Phase I assesses safety and dose in small numbers; Phase II looks for preliminary efficacy; Phase III is a large comparison against existing options to support approval; and Phase IV monitors performance after a product is on the market.
Why must clinical trials be registered?
Prospective registration creates a public record of a trial’s aims and design before results are known. It deters selective reporting, reduces publication bias, and is required by the ICMJE as a condition of publishing the results.
What is the difference between a clinical trial and clinical research?
A clinical trial is one type of clinical research in which an intervention is tested under a protocol. Clinical research is the broader field, which also includes observational studies that do not assign an intervention.
What governs the ethics of a clinical trial?
Independent ethics committees, informed consent, the Declaration of Helsinki and Good Clinical Practice together govern trial ethics. See our Good Clinical Practice explainer.
Leave a Reply