Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

A small girl with curly hair is sitting on a man's shoulders. She is using her hands to form binoculars to look through.
Image credit: Edi Libedinsky via

I know one when I see it

Everyone knows what a clinical trial is, right? A bit like a good surgeon, a fine wine, or a good whisky, we know a high-quality research study when we see one. It may surprise some to find that there might be difficulty in defining a clinical trial. I imagine those people haven’t spent too long in the company of academics. No topic too miniscule, or focus too narrow to preclude a difference in opinions, find methodological challenges, or for us to find distinctions. Professional integrity is at stake for us academics.

Being more serious, clinical trials are simple to spot and easy to define right up to the point when one tries to give the exact definition. A natural place to start would be to look at what others, including some pioneers, have concluded.


Early definitions

It is unclear when the term “clinical trial” was first coined, though it was certainly in use in the 1930s. It seems to have originated from those involved with the Medical Research Council (MRC) in the UK. The MRC set up a “Therapeutic Trials Committee” in 1931 with studies coming out of the MRC later described as “clinical trials”. The term “trial” was used to refer to studies in the 1920s and indeed goes back even further (e.g. Heiberg, 1897). The term “clinical trial” over time became the predominant one, encompassing other terms like “therapeutic trial” and some “field trials”. The latter term was sometimes used for vaccine trials as the target population was not patients.

The late, great Bradford Hill of MRC Streptomycin trial fame does not give a definition in his 1951 and 1952 papers on clinical trials (Hill, 1951, 1952), despite both publications being titled “The clinical trial”. Always a thoughtful and careful writer, he presumably considered it fairly obvious at the time, at least if one were to read the paper. These papers, particularly the 1952 one, worked through the basics of how a trial should be designed, conducted, and reported. From the 1952 paper, one can surmise what he had in mind, at least regarding what he thought clinical trials should be. Here he refers to “controlled clinical trials”, so a clinical trial may not be controlled, although preferably it would be, and that control would ideally be a “random” one. Similarly, a trial involves patients and at least one treatment being received by them, has measurements that lead to "results", and requires "reporting".

“Returning to the aims of the trial, it is essential that these should be laid down in every detail at the outset” (Hill, 1951). So the aims of clinical trials could vary in a non-trivial way, which is both of interest to note and also challenging regarding our definition.

To some of the leading advocates of clinical trials, they were pretty much synonyms for a time with RCTs. Indeed one might view the RCT as the “ultimate” or optimal RCT design, with the caveat that one has to be “ready” to conduct one (i.e., know enough for example about the dose). However, this definition of clinical trials is widely recognized these days as unhelpfully narrow, even if many might still agree that the RCT is the bee's knees of clinical trials (all other things being equal).


More recent definitions

So what have others said more recently?

In their book Modern Epidemiology, Rothman and Greenland, while not stating it as a formal definition, say “ A clinical trial is an experiment with patients as subjects”. Which restricts clinical trials to studies involving patients and also begs the question in this content: what might constitute an experiment? Helpfully, Rothman and Greenland defined an experiment as “any test or trial” (page 69, Rothman and Greenland), which is not so helpful in defining a clinical trial, and, thankfully for this purpose, state that an experiment is “a set of observations, conducted under controlled circumstances, in which the scientist manipulates the conditions to ascertain what effect such manipulation has on the observation” (page 76, Rothman and Greenland).

Woodward (2014; page 257), suggests a clinical trial is synonymous with an “intervention study” and is “an experiment applied to existing patients, in order to decide upon an appropriate therapy or to those presently free of symptoms, in order to decide upon an appropriate preventive strategy”.

Machin and Fayers (2010; page 2) defined a clinical trial (based on Day, 2007) as “Any systematic study of the effects of a treatment in human subjects”. So human subjects (not limited to patients) and focused on the study of a treatment. Pretty broad but perhaps not covering everything described as a clinical trial in practice.

In his seminal book, Pocock (1983, page 1) suggested that a clinical trial could be “any form of planned experiment which involves patients and is designed to elucidate the most appropriate experiment of future patients with a given medical condition”. So a planned experience, involving patients, and about learning how to treat patients with a specific condition. This definition covers most but not all clinical studies, if one is being picky, such as some that include multiple conditions and those on healthy volunteers.

Friedman and colleagues (2015) defined a clinical trial as “a prospective study comparing the effects and value of intervention(s) against a control in human beings.” This definition is similar to Pocock’s quoted above, broader in that it uses terms like “intervention(s)”, but narrower as it implies there needs to be a control group.


What about “authoritative” bodies

The FDA's website defines a clinical trial as “Clinical trials are research studies in which people volunteer to help find answers to specific health questions” (US Food & Drug Administration, 2023). So it is a research study, involving volunteers, that addresses health questions. However, that does not seem to exclude a cross-sectional questionnaire study nor other studies that one might not wish to fall under the term clinical trial.

The NIH defines a clinical trial as a “Research study in which one or more human subjects are prospectively assigned to one or more interventions (which may include placebo or other control) to evaluate the effects of those interventions on health-related biomedical or behavioral outcomes” (National Institutes of Health Research, 2014).

The UK’S MHRA focuses on clinical trials of investigational medicinal products (CTIMPs) and are concerned with the definition under the relevant legislation, which was defined in common with the other (at the time) EU members states (Medicines and Healthcare products Regulatory Agency, 2023). The MHRA’s focus is not surprising given that it has authorisation authority only for those trials that fall with its remit: CTIMPS and some medical device studies. The corresponding 2004 legislation definition is:

"clinical trial" means "any investigation in human subjects, other than a non-interventional trial, intended:

(a) to discover or verify the clinical, pharmacological or other pharmacodynamic effects of one or more medicinal products,

(b ) to identify any adverse reactions to one or more such products, or

(c) to study absorption, distribution, metabolism and excretion of one or more such products, with the object of ascertaining the safety or efficacy of those products." (UK Government Legislation, 2004).

It’s interesting to note that this definition is expressed mostly in pharmacological terms, although it does include “To discover or verify/compare its clinical effects” as the stated objective of such a study, which is a much broader definition that could be applied much more widely. Similarly, “human subjects” are the subjects of the study and the study is an “interventional trial” as it explicitly excludes “non-interventional trials”.

While legal definitions of a clinical trial vary somewhat, existing legal definitions focus on medicines (e.g., UK Government Legislation, 2012). In the 1968 Medicines Act in the UK, clinical trials were similarly defined in terms of evaluating medicines, and it was refers to "an investigation or series of investigation". In this sense one might be tempted, wrongly in my view, to think that clinical trials are only studies that evaluate drugs (strictly speaking IMPs or investigation drugs to my American friends, although interestingly the legislation in the US does not define a clinical trial).


My definition

In my recently published book (Cook, 2023), I created my own definition: “... a clinical trial can be described as a planned scientific study involving human participants which is conducted to learn about the safety and effect (e.g., ‘efficacy’) of a medical treatment or related care” (page 1). This was in part developed after looking at a variety of definitions (though not exactly those stated above) and also thinking about common usage. Key points for me are:

  • It is clearly a scientific or research study.
  • It involves human participants, although participants can be via another's consent (e.g. a parent).
  • It looks at the safety and/or effect of something given or done to people directly or indirectly (perhaps very indirectly).
  • It is about a medical treatment or related care (which is by no means restricted to drugs and interpreted inclusively).

As with all of the previous definitions, one might critique. You may have a better one.

What is clear is that the term clinical trial is used to cover a wide variety of studies from first in human studies of a new drug to large patient-based studies of a wide variety of treatments. “Treatments” refers to both those in the traditional sense, anything potentially therapeutic (e.g. acupuncture), and even related care strategies (e.g. fast track surgery).

So for me, the term "clinical trial" needs to be a broad concept covering exploratory, early phase, late phase, and those that evaluate treatments already available in medical practice. Although most RCTs are clinical trials, not all clinical trials are RCTs. So it should not be thought of as a specific design, though a RCT might be considered the ideal design (with some caveats). The definition of clinical trial also needs to allow for interventions other than drugs or, to use the legalised form, investigative medical products. Otherwise we have the absurd scenario that whether a study is a clinical trial depends upon whether an intervention is viewed as an IMP or not, which can lead to odd divisions either way. Instead, surely all trials of medical treatments or interventions (in the broad sense as long as medically related) could be thought as a clinical trial. The final division in clinical trials is medical versus general health, which can include “public health” interventions such as an RCT of health interventions for school children. I would readily consider a trial of vaccine effectiveness (does it work in practice) as a clinical trial. I would like generally include RCTs conducted on health volunteers recruited from the community. But perhaps not all public health RCTs are “clinical trials”. So fine lines, and perhaps there is no perfect definition. Indeed, one might argue that is very fitting as the name follows the nature in that things are more complicated in practice than might initially appear to be the case. Few who have worked in clinical trials, I think, would dissent from that sentiment.



Cook JA. An Introduction to Clinical Trials. 1st Edition, OUP, 2023. Available from: An Introduction to Clinical Trials | Oxford Academic (

Day S. Dictionary of Clinical Trials, 2nd edition, John Wiley & Son, 2007.

Friedman LM, Furberg CD, deMets DL, Reboussin DM, Granger CB. Fundamentals of Clinical Trials. Fifth Edition, 2015. Springer.

Heiberg P. Studier over den statistike Undersogelsesetode som Hjaelpemiddlel ved terapeutiske Undersogelser. 1897.  The James Lind Library.  Available from:

Hill AB. The Clinical Trial. British Medical Bulletin. 1951;7(404):278-282. doi: 10.1093/oxfordjournals.bmb.a073919.

Hill AB. N Engl J Med. 1952;247(4):113-9. doi: 10.1056/NEJM195207242470401.

Machin and Fayers PM. Randomized Clinical Trials: Design, Practice and Report. 1st edition, Wiley-Blackwell, 2010.

Medical Research Council Therapeutic Trials Committee. The serum treatment of lobar pneumonia. BMJ 1934;i:241­5.

Medicines and Healthcare products Regulatory Agency. Clinical trials for medicines: apply for authorisation in the UK. Last updated: 26 June 2023. Available from:

National Insitutes of Health Research. Notice of Revised NIH Definition of “Clinical Trial”. 23 October 2014. Available from:

Pocock SJ. Clinical trials: a practical approach. 1st Edition, John Wiley & Sons Ltd, 1983.

Rothman and Greenland, Modern Epidemiology, 2nd edition, Lippincott-Raven Publications, 1998.

UK Government Legislation. The Medicines Act 1968. UK Statutory Instruments 1968. Available from: Medicines Act 1968 (

UK Government Legislation. The Medicines for Human Use (Clinical Trials) Regulations 2004. UK Statutory Instruments 2004 No. 1031 Part 1 Regulation 2. Available from:

UK Government Legislation. The Human Medicines Regulations 2012. UK Statutory Instruments 2012 No. 1916. Available from

US Food & Drug Administration (FDA). Basics about clinical trials. Last updated 08 May 2023. Available from:

Woodward M. Epidemiology: Study Design and Data Analysis. 3rd Edition, CRC, 2014.

Written by

  • Jonathan Cook
    Jonathan Cook

    Professor of Clinical Trials & Medical Statistics/OCTRU Deputy Director

Latest Haphazard Thoughts

To randomise or to randomize?

A look at the use of randomis(z)e and randomis(z)ation when discussing RCTs.