target trial emulation (TTE) is a research method that uses real-world observational data to mimic a randomized controlled trial (RCT). It controls confounding factors through standardized research design and causal inference methods, so that the research results reach or approach the reliability level of RCT. In the TTE framework, propensity score (PS) is the main method to control confounding bias, and its role is similar to randomized allocation in RCT, making the baseline characteristics between groups reach equilibrium. PS refers to the covariate at a given observedXThe conditional probability of an individual being assigned to a particular treatment group under the condition of, can be expressed as:ei=P（Ti=1|Xi), whereinTi=1 denotes an individualiAre assigned to a processing group,XA vector composed of covariates (e.g. age, gender, etc.). The obtained covariate information (i.e., potential confounders) can usually be used to predict the allocation probability of the intervention based on Logistic regression models or other models, and the PS can be constructed for each study subject. However, PS itself does not directly control confounding bias, and it is necessary to further achieve inter-group equilibrium through the following methods:
