A clinical trial is a structured research study involving human participants that prospectively assigns one or more interventions to evaluate their effects on health-related outcomes. Interventions can include medicines, vaccines, medical devices, surgery, diagnostic approaches, behavioural interventions, dietary strategies or new ways of delivering healthcare. The defining feature is prospective assignment rather than simply observing what people independently choose to do. NIH's formal definition specifies human participants, prospective assignment to an intervention and evaluation of health-related biomedical or behavioural outcomes.
Clinical trials are conducted to answer focused questions about safety, efficacy, effectiveness, dosage, side effects, prevention, diagnosis or quality of life. Treatment trials may compare a new medicine with placebo or an existing standard treatment. Prevention trials may test vaccines or interventions intended to reduce disease risk. Diagnostic trials can compare new tests with an accepted reference standard. Behavioural trials can evaluate interventions designed to alter behaviours relevant to health. The design therefore depends on the research objective rather than all clinical trials following an identical structure.
A clinical trial begins with a protocol, which is the formal plan specifying the research question and how the study will be conducted. It typically defines eligibility criteria, treatment groups, intervention doses, schedules of visits, outcome measures, safety monitoring, statistical analysis and procedures for handling adverse events. The protocol is designed before the main data collection so that researchers cannot easily alter the study's fundamental methods after seeing results in ways that might bias the findings. Participants must receive information about relevant risks and benefits and provide informed consent before participation.
Participants may be assigned to different groups using randomisation. Randomisation allocates treatments by chance and helps prevent systematic differences between groups from influencing the comparison. Trials may also use control groups receiving placebo, standard treatment or another comparator. The comparator provides a benchmark against which outcomes can be interpreted. Without an appropriate comparator, improvement after an intervention may be difficult to distinguish from natural recovery, regression to the mean, placebo responses or other changes occurring over time.
Blinding or masking is another important trial-design feature. In a blinded study, participants, researchers, outcome assessors or other personnel may be unaware of treatment allocation. The purpose is to reduce bias arising from expectations or differential treatment. Double-blind trials are often particularly useful for subjective outcomes, although blinding can be impossible for some interventions, such as surgery or visibly different behavioural programmes. When blinding cannot be achieved, researchers can still reduce bias through objective outcomes, independent outcome assessment, predefined protocols and appropriate statistical methods.
Clinical trials must balance scientific objectives with participant safety. Ethical review evaluates whether the risks are reasonable in relation to potential benefits and scientific value. Participants are monitored for adverse events, and predefined safety procedures can lead to dose changes, protocol amendments or trial termination. Trials involving serious risks may have independent monitoring structures. Participants can generally withdraw from research, subject to the details of the study and applicable regulations. These protections are necessary because the scientific value of a study does not override the rights and welfare of individual participants.
Clinical trials generate several types of information. Researchers may measure efficacy under controlled conditions, safety and adverse events, pharmacokinetics, treatment adherence, quality of life, mortality, disease progression or other outcomes. The choice of outcome matters because a statistically significant change in a surrogate laboratory marker does not necessarily translate into meaningful improvement in how a patient feels, functions or survives. Good trials therefore prespecify clinically relevant primary and secondary outcomes and analyse them according to an established statistical plan.
Clinical trials form one part of the wider medical research process. Candidate interventions may first be investigated through laboratory and preclinical research, then assessed in increasingly informative human studies. Positive trial results do not automatically mean that an intervention is suitable for every patient because trial populations may differ from routine clinical populations and follow-up may be limited. Later studies, post-marketing surveillance and real-world evidence can reveal uncommon adverse effects and longer-term outcomes. NIH describes clinical trials as central to determining whether approaches to prevention, diagnosis or treatment are safe and effective.
