Study Design in Medical Research

All studies conducted in medicine can be broadly classified into descriptive and analytical studies.

Descriptive Studies

Descriptive studies are not designed to determine relationships between different factors. Instead, they aim to describe what is happening within a population, such as the prevalence of a disease.

Descriptive studies include:

• surveys;
• qualitative research methods, including case reports;
• case series, which help determine the prevalence of specific factors and, consequently, the scale of a given problem.

Analytical Studies

Analytical studies are designed to identify relationships between different influencing factors, for example, the effect of a specific treatment method on patient outcomes.

To assess an effect, results must be compared between:

• a control group;
• an experimental group.

Types of Analytical Studies

Analytical studies can be divided into:

• experimental studies (with active intervention by the researcher);
• analytical observational studies (based on passive observation).

In analytical observational studies, the researcher measures the level of exposure or treatment effect within a group of patients.

Analytical Observational Methods

Analytical observational methods include:

• cohort studies;
• case–control studies;
• cross-sectional studies.

Cohort Studies

A cohort study is an observational study in which a defined group of individuals (a cohort) is followed over time. Outcomes are compared between subgroups that were exposed or not exposed to a specific treatment or factor. These studies may be retrospective or prospective.

Case–Control Studies

Case–control studies compare individuals with a specific disease or outcome (cases) to individuals from the same population without that disease or outcome (controls) in order to identify associations between outcomes and prior exposure to risk factors.

Cross-Sectional Studies

Cross-sectional studies examine a defined population at a single point in time and provide information on disease prevalence and characteristics.

Clinical experimental studies are considered the most informative and reliable.

Clinical Experimental Studies

In experimental studies, the researcher determines the intervention to which patients in the control and experimental groups will be exposed.

Randomized controlled trials (RCTs) are analogous to experiments in other scientific fields. Patients are assigned to two or more groups and receive specific interventions under carefully controlled conditions.

The most commonly used designs include:

• parallel group design;
• crossover design.

Parallel Group Design

In parallel group clinical trials, patients in two or more groups receive different therapies. To ensure statistical validity and minimize systematic bias, patients are allocated to groups using randomization.

A typical model can be illustrated as follows:

Screening → Enrollment → Run-in period → Baseline assessment → Randomization →
Treatment A → Outcomes A
Treatment B → Outcomes B

where A and B represent different treatment methods.

Parallel group trials:

• are expensive and time-consuming;
• require a large number of participants (commonly 30–50 per group);
• are considered the most objective for evaluating treatment effectiveness and drawing conclusions.

Crossover Design

Unlike parallel group designs, crossover studies allow the evaluation of treatment effects within the same group of patients. Patients are randomized to receive treatments in different sequences.

A washout period between treatment phases is essential to allow outcomes to return to baseline and to eliminate residual effects of the previous treatment.

A typical design includes:

Screening → Run-in period → Baseline assessment → Randomization →
Treatment A (Group 1) / Treatment B (Group 2) → Washout →
Treatment B (Group 1) / Treatment A (Group 2)

A simplified version may include a single group receiving a control treatment, followed by a washout period, then the experimental treatment, another washout period, and a return to the control treatment.

Example of a Crossover Study

A similar design was applied in a study of conductive education [1].
Initially, 143 children were assessed, and 67 were selected for further participation. After the first assessment, all children received standard special education for four months, including two 30-minute therapy sessions per week using Vojta or Bobath methods.

Subsequently, over nine months, standard special education was combined with courses of conductive education. All children completed three one-month inpatient courses of conductive education, with intervals of standard therapy. A final assessment was conducted after an additional four months of standard treatment.

Limitations of Crossover Designs

Crossover designs are more economical than parallel group studies because they require fewer participants. However, they may present challenges in interpreting results due to:

• carryover effects between treatments;
• difficulty distinguishing sequential treatment effects from individual intervention effects.

References

1. Blank R, Kries R, Hesse S, Voss H. Conductive Education for Children with Cerebral Palsy on Hand Motor Function Relevant to Activities of Daily Living. Archives of Physical Medicine and Rehabilitation. 2008;89:251–259.

About the author

Taras Voloshyn

Neurologist at the Kozyavkin International Rehabilitation Clinic since 2009. Conference participant, author of publications, specialized in the assessment and rehabilitation of patients with cerebral palsy.