This section provides an explanation of the common terms you might encounter during research. If there is a term you think should be added to this list, please email web@imperial.nhs.uk.
Arm
A group of patients receiving a particular treatment (or placebo) in a clinical trial
Blind
This term applies to trials where a new treatment is compared against another treatment or a placebo (usually phase III trials). A trial is ‘single-blind’ if either the participants or the researchers do not know which option the participants are receiving, and ‘double-blind’ if both the participants and the researchers do not know which option the participants are receiving. This makes the results of the trial more objective and therefore more reliable
Cohort
A group of individuals with some characteristics in common, e.g. a group of people born within the same period would be referred to as a birth cohort
Contraindication
A specific circumstance when the use of a treatment could be harmful, e.g. aspirin allergy is a contraindication to taking aspirin
Control/control group
An alternative to a new treatment that is used as a comparison. Controls are usually either an existing treatment, a placebo or no treatment
Crossover trial
A trial that involves patients receiving one treatment for a certain amount of time then swapping to an alternative treatment for a certain amount of time
Double-blind
See ‘blind’
Eligibility criteria
A description of people that can (inclusion criteria) or cannot (exclusion criteria) take part in a trial
Informed consent
The process of learning what is involved in a clinical trial and then agreeing to take part
Longitudinal study
A study that follows a group of patients over a period of time
Meta-analysis
A review of the results of a large number of trials on a similar subject. A meta-analysis can be a particularly powerful research tool
Multi-centre trial
A trial being carried out at more than one location
Observational study
A study where the researcher is not directly controlling the experiment but instead observing behaviour or outcomes. No attempt is made to affect the outcome, e.g. no treatment is given
Peer review
A review of a trial’s results by a group of independent experts
Phase I trial
Phase I trials focus on the safety of a new medical treatment. They involve only a small number of people, who are usually healthy volunteers. The trials look at several measures of safety, such as side-effects of the treatment
Phase II trial
Phase II trials focus on whether a new medical treatment is effective. They generally involve a larger group of volunteers than phase I trials. The volunteers are usually patients who have the condition the treatment is intended for. Phase II trials also look at safety
Phase III trial
Phase III trials focus on the specific benefits of a new medical treatment by comparing it with a similar treatment or a placebo (a ‘dummy’ treatment). Patients who receive a placebo should still also receive the best standard care. Phase III trials last longer than phase I or II trials (usually a year or more) and involve more people (often several thousand treated across different hospitals in different countries)
Placebo
A ‘dummy’ treatment that resembles a medical treatment but is intended to have no physical effect on a participant. A new treatment is often compared against a placebo to get more reliable evidence about its effectiveness. Patients who receive a placebo should still also receive the best standard care
Prospective study
Research in which a group of participants is identified, and then studied from that point forward in time. The opposite is a retrospective study
Randomised trial
In a randomised trial, participants are allocated to receive one type of treatment or another by a random process, usually using a computer. This helps ensure the results are objective and unbiased
Retrospective study
Research in which a group of participants is identified, and then studied from that point backward in time, usually via their medical records and interviews. The opposite is a prospective study
Single-blind
See ‘blind’
