This article presents propensity score matching (PSM) - modern statistical method to control for confounding, which threatens the validity of associations in observational studies. The efficiency of PSM has been demonstrated in several international studies. The increasing use of PSM by the research community is reflected by steady growth of the number of publications with this method in the PubMed database over time. This article presents the theoretical basis of PSM and its practical application using Stata software. In the practical part of the article, detailed step by step algorithms of various PSM methods are presented allowing researchers to conduct statistical analysis of their own data and interpret results.

The authors presents a propensity score matching (PSM) technique - an effective method to control the effect of confounding factors in observational studies. PSM has been shown to be as efficient as linear regression analysis, but can be performed using smaller samples. This article presents the basic principles of PSM and its practical application using STATA 13 software for the studies with continuous dependent variable.