[Application of regression discontinuity design in epidemiological research].

ABSTRACT

Regression discontinuity design (RDD) is a rigorous quasi-experimental method. Based on the randomness of the distribution of samples around the critical value, the data on both sides are used for regression respectively, so as to avoid the interference of potential confounding factors and provide strong evidence for the inference of causal correlation. This study introduces the RDD and its research progress, and expounds its application in detail combined with cases, in order to provide reference for the application of RDD in epidemiological research.