A clinical study is a scientific study involving people that is conducted to evaluate the efficacy and safety of a new drug or expand the indications for the use of an already known drug. Clinical studies can also examine the efficacy and safety of new invasive (including surgical) and non-invasive treatments and diagnostics.
Clinical studies around the world are an integral stage in drug development, which precedes its registration and extensive medical use. In clinical trials, a new drug is being studied to obtain data on its efficacy and safety. Based on these data, the authorized health authority decides whether to register the drug or refuse to register. A drug that has not undergone clinical studies can not be registered and put on the market. When developing a new drug, clinical trials can not be avoided, since the extrapolation of research results in animals and on biological models per person is possible only in general, and sometimes impossible at all. For example, the pharmacokinetics (how a drug enters the blood, is distributed in the body and is removed from it) in humans differs even from the pharmacokinetics in primates. However, the analysis of preclinical studies is very important for assessing the likelihood of development and the nature of side effects, calculating the starting dose for studying the properties of the drug in humans. Clinical studies can only be initiated after encouraging results have been obtained in preclinical studies (studies on biological models and laboratory animals), as well as the approval of the ethical committee and the positive decision of the health authority of the country where the study is planned
The design of a clinical trial is a plan for its conduct. The design of a specific clinical study depends on the goals pursued by the study. Consider three common design options:
- single group design
- clinical research in parallel groups
- clinical research in the “crossover group design”
Clinical study in one group
(single group design)
In the study in one group, all subjects receive the same experimental treatment. This model of research is aimed at comparing the results of treatment with the baseline condition. Thus, subjects are not randomized to treatment groups.
The model of clinical research in one group can be illustrated as follows:
Screening – Inclusion – Initial state – Treatment – Outcomes
The model of one group can be used in the first phase of research. Models of studies in one group are usually not used in phase III clinical trials. The main drawback of the model of research in one group is the lack of a comparison group. The effects of experimental treatment can not be differentiated from the effects of other variables.
Clinical research in parallel groups
(parallel group design)
When conducting clinical trials in parallel groups, subjects in two or more groups receive different therapies. To achieve statistical reliability (to exclude a systematic error), the subjects are distributed in groups by the method of random distribution (randomization).
The model of clinical research in parallel groups can be illustrated as follows:
Screening – Inclusion – Preparatory period – Initial state – Randomization –
Treatment a – Outcome a
Treatment b – Outcomes b
Where a, b are different drugs or different doses or placebo
Clinical studies in the design of parallel groups are expensive, time-consuming and require a large number of subjects (with a low incidence of events considered). However, clinical studies in parallel groups are the most objective in determining the effectiveness of treatment and are accurate in the formulation of findings. Most clinical trials, therefore, are conducted in the design of parallel groups.
Sometimes studies in parallel groups can be used in two versions – a factorial and heterogeneous model. Factorial design is a design on the basis of several (more than 2) parallel groups. Such studies are conducted when it is necessary to study a combination of different drugs (or different doses of a single drug).
The factorial model of clinical research can be illustrated as follows:
Screening – Inclusion – Preparatory period – Initial state – Randomization –
Treatment a – Outcome a
Treatment b – Outcomes b
Treatment with – Outcomes with
Treatment in – Exodus in
Where a, b, c, d are different drugs or different doses or placebo
The factorial model is useful in evaluating combined medicines.
The drawback of the factorial model is the need to attract a large number of subjects and as a result – increase the cost of research.
Inhomogeneous (interruptible) model of “termination of therapy” (Withdrawal (Discontinuation) Design)
The heterogeneous model is a variant of studies in parallel groups, where all subjects are first given experimental treatment, then patients are randomly assigned to receive the appropriate reactions in groups using double blinded or placebo blind blind technology. This model is usually used to evaluate the effectiveness of experimental treatment by stopping the drug immediately after the appearance of the reaction and recording the recurrence or remission. In Fig. 5 shows the scheme of the inhomogeneous model of research.
Screening – Inclusion – Experimental treatment – Response to treatment – Randomization of responders to treatment – Treatment or Placebo
A non-uniform model of research is especially effective for the evaluation of drugs intended for the therapy of intractable diseases. In carrying out such studies, only a small percentage of the subjects demonstrate the response to treatment.
During the treatment period, responses are identified, and the randomization phase according to the inhomogeneous model is used to demonstrate that the given reaction is real, and not a reaction to placebo. In addition, heterogeneous models are used to study relapses.
The disadvantages of heterogeneous models are:
a large number of subjects who initially receive treatment to detect responses
significant duration of the study
The preparatory period should last long enough for the patient’s condition to stabilize and the drug effect to be more clearly revealed. It should be noted that the percentage of subjects excluded from these studies may be high.
Ethical standards require careful consideration of the application of this model of research because it may be necessary to exclude from therapy that medicine that brings relief to patients. Strict monitoring and a clear definition of endpoint indicators are of paramount importance.
“Cross-over” model (Crossover Design)
Unlike research plans in parallel groups, “cross” models allow to estimate the effects of both the studied drugs and comparative courses of treatment on the same subjects. Subjects are randomized to groups in which the same course of treatment is given, but with varying consistency. As a rule, a “washout” period is needed between the courses in order for the patients’ parameters to return to the initial ones, and also to exclude the undesirable effect of the residual phenomena of the previous treatment on the effects of the subsequent one. A “washout” period is not necessary if the individual’s reactions are limited to a comparison at the end of each course, and the treatment period lasts long enough. In some “cross” models, a preliminary “crossing” is used, which means that patients who are excluded from studies at the treatment stage can be transferred to alternative treatment groups earlier than planned.
In Fig. 6 is a diagram of a typical “cross-sectional” research model.
Screening – Preparatory period – Status monitoring – Randomization – Treatment A in group 1 and Treatment B in group 2 – Washing period – Treatment B in group 1 and Treatment A in group 2
“Cross-over” models are commonly used to study pharmacokinetics and pharmacodynamics when the task of controlling variability in a population of subjects is set. In addition, it is fair to assume that the effects of the first course do not affect the second in pharmacokinetic and pharmacodynamic studies with a sufficient “wash-out” period.
“Cross-over” models are more economical compared to parallel group models, since in this case fewer subjects are required. However, sometimes there are difficulties in interpreting the results. The effects of one therapy can be mixed with subsequent effects. It can be difficult to distinguish the effects of sequential treatment from the effects of individual courses. In clinical trials, the “cross-over” model usually requires more time than parallel group studies, because each patient undergoes no less than two treatment periods plus a “washing-out” period. This model also requires obtaining more characteristics for each patient.
If the clinical conditions are relatively constant throughout the study period, the “cross” model is effective and reliable.The relatively low requirements for sample size make “cross-over” models useful in early clinical development in order to facilitate decision making on more voluminous models of parallel studies. Since all subjects receive the study medication, the “cross” studies are also effective for safety assessment.
Great importance in the conduct of clinical research is played by psychological, or so-called subjective factors. For example, a patient’s knowledge that he is receiving active drug therapy may affect the safety and efficacy of therapy. A medical researcher who is convinced of the advantages of one of the drugs compared can involuntarily interpret in his favor improvements in the state of health of patients or try to appoint a patient with a more severe disease that treatment which he considers more effective. To minimize the influence of subjective factors, blind research is used.
A study in which the patient does not know, and the researcher knows what treatment the patient receives is called a simple blind person. If neither the patient nor the researcher knows about the prescribed treatment, such a study is called double-blind. Blind studies can minimize the possibility of intentional distortion, and unintentional – distribute between groups in equal proportions.
