What is a Clinical Trial?
The most basic question in medicine is: what should I do for this patient? Evidence-based medicine uses the results of scientific studies to give us the answer. The best answers usually come from randomized clinical trials – the focus of the ISN-ACT Toolkit.
Clinical trials (also known as experimental or interventional studies) must be distinguished from the other main type of clinical study – observational studies. The key characteristics of a clinical trial are that an INTERVENTION is deliberately assigned to participants by the researchers. Often the intervention is assigned to participants randomly – in a process known as RANDOMIZATION. This process is the best for accounting for CONFOUNDING factors (both known and unknown) and so is key to minimizing BIAS. For a more in-depth perspective of the types of clinical trials see Trial design.
Non-randomized studies (eg. cohort and case-control studies, case series and case reports, and ecological or community studies) are essential to develop new hypotheses in medicine. Yet it is important to recognize that it can be difficult to be sure of causation from such observational studies as the treatment (or other risk factor, exposure, or variable of interest) that the study participants experience is not assigned in an even-handed way by randomization. In an observational study, the factors that govern why one patient receives a treatment and another does not, or why a particular patient was exposed to a risk factor or had their care managed in a certain way, may also influence the outcome of the patient’s disease. Some of these factors can be identified and controlled for using statistical or design methods, but there may always be factors that remain unknown or unmeasured. For this reason, where feasible and ethical, one should always consider incorporating randomization in study design.
Remember that randomized clinical trials do not just have to test treatments, but should be considered for any process or intervention – including such diverse things as counselling, patient education, communication services, diagnostic tests, and even clinician-facing processes such as automated alerts.
Clinical trials of drugs or other therapies are usually classified by ‘phase’. This divides them into four types of study depending on the type of question that is being answered.
Phase I trials are focused on dosage and safety. They typically have two main questions: What dose should be given (pharmacokinetics and pharmacodynamics)? What side effects can occur (toxicity)? They are usually conducted with small samples of healthy volunteers.
Phase II trials aim to test if the treatment works. They have one main question: Does this treatment have the intended effect in real patients (EFFICACY)? In this phase, safety continues to be evaluated. Usually, these studies are conducted in a relatively small sample of highly selected patients, in studies often referred to as “proof-of-principle”.
Phase III trials aim to confirm the efficacy (and safety) of treatments by testing them in a large and more diverse patient population. These trials differ from Phase II studies by aiming to confirm efficacy in a GENERALIZABLE patient population. They are typically required for approval of a treatment by regulatory bodies (such as the FDA).
Phase IV trials describe clinical trials occurring after a drug has been approved for marketing. These trials gather additional information about a drug’s safety, efficacy, or optimal use. Phase IV trials may aim to enroll a diverse group of participants in order to increase generalizability. An important function of Phase IV trials is to compare two different treatments for the same condition – so called COMPARATIVE EFFICACY studies. In this case, both trial arms receive active treatment. In addition, many Phase IV trials aim to determine EFFECTIVENESS, by enrolling a diverse group of participants and ensuring that their clinical treatment is as close as possible to that received in usual practice. Such studies often also consider the costs and benefits of different treatments to enable more efficient use of health resources.
In practice, many clinical trials do not fit neatly into a single phase. What is most important for designing your own clinical trial is to clearly define the clinical question that you want to answer. For assistance with these first steps see: Study stage 1: Design and development.