Studies and research design in medicine

Abstract

Adequate design is an essential condition for conducting a successful study. This review describes the most common types of research design in medicine. We discuss the differences between different types of observational and interventional studies, their advantages and limitations providing examples for each study design. The concept of bias and its potential sources in different studies are covered. We suggest the most suitable approaches to study design for different research objectives and outline approaches to data presentation. During the last decades, several guidelines for conducting and reporting different types of research were proposed and they are also covered in this manuscript.

Statistical hypothesis testing: general approach in medical research

“…Такие исследования получили названия гипотезообразующих (о различных подходах к оптимальному выбору дизайна исследований для различных целей -см. [5]). Безусловно, если удалось зафиксировать эффект, в пилотных исследованиях можно оценить его наблюдаемый или гипотетический размер.…”

Basic principles of descriptive statistics in medical research

“…The choice of variables depends on the study objectives. Exposures and outcomes are also variables from the statistical point of view (more on exposures and in outcomes in different study designs in the previous tutorial) [3]. The examples of variables include sex, age, concentration of a biomarker, gravidity (number of pregnancies), tumor grade, history of smoking etc.…”

“…In other words, in every five patients, one will develop MI, and four will not. Odds are rarely reported per se, but they are often used to assess the relationship between two binary variables, for example the presence (or absence) of the risk factor and outcome, by calculating odds ratio (more on that in the previous) [3].…”