How is the study design defined

Study design

English: study design

1 definition

Under the Study design one understands in medicine the totality of the procedures in the context of a study. A good study design tries to prevent influences that can falsify the result of a test procedure.

2 methodological standards

The specific procedures in a study essentially depend on the objective, the type of treatment (s) to be tested and the study indication.

Non-interventional studies are usually uncontrolled studies because they are conducted without a control group.

Prospective, interventional clinical studies to demonstrate efficacy and safety are usually carried out as randomized controlled studies.

3 procedures

3.1 Type of control

In controlled studies, the treatment to be tested (drug or other intervention) is compared against one or more control treatment (s).

In a placebo-controlled study, the control group receives a dummy drug (placebo).

In equivalence or non-inferiority studies, the control group receives a comparator drug (usually the current standard therapy). Depending on the question, non-drug treatment methods can also be used as controls. Untreated patients can also serve as controls.

When composing a control group, it must be ensured that the persons in the control groups roughly match in terms of age, gender, etc. (see section randomization).

3.2 Cross-over / parallel groups

A controlled clinical study can be carried out either as a cross-over study or in a parallel group design.

3.2.1 Cross-over

In a cross-over study, the same test persons are given the test product and the control product one after the other (e.g. bioequivalence studies). Through the randomization, each participant is assigned the sequence of the treatments at random. A wash-out phase must then be observed between the treatments, which is determined on a study-specific basis depending on the pharmacokinetic properties of the drugs.

3.2.2 Parallel groups

Therapy studies are usually carried out in parallel groups. In addition to the experimental group, which receives the test preparation, one or more control groups are treated in the study. The assignment to the respective treatment group is random (randomized).

3.3 Blinding

A controlled study can be open, single-blind, or double-blind.

  • Open means that the treatment is open to everyone involved (investigator and patient).
  • In most cases, studies are carried out as double-blind studies in which neither the doctor nor the patient is informed about the treatment. This serves to prevent the effect of self-suggestion and also to control the Rosenthal effect. The double-dummy method can also be used to compare two different dosage forms (e.g. tablet versus ointment) as a double-blind study.
  • At easy In blinded studies, only one of the participants is blinded, either the investigator (investigator-blind) or the patient (subject-blind). The reason for a simply blinded study can be the type of treatment, for example if it cannot be avoided that the patient can recognize it when taking it (e.g. lozenges by the taste).

3.4 Randomization

In order to increase the representativeness and comparability of the samples, a randomization is carried out, i.e. the individual subjects are randomly distributed to the treatment groups.

The randomization ensures that the comparison groups differ as little as possible in terms of their characteristics at the beginning of treatment (structural equality of the initial conditions). This relates both to general characteristics, such as the age distribution, as well as to the characteristics of the disease that is examined in the study.

3.5 parallelization

The process of parallelization also serves to increase the representativeness and comparability of samples. In this case, a person from the experimental group is assigned to a person from the control group, whereby both persons have the same characteristics (age, gender or personal characteristics such as extroversion, etc.). This process is mainly used for small samples.

see also:Clinical Trial Directory

4 sources