Fig. 3

Associations between AF and outcomes in the crude analysis, multivariable analysis, and propensity score analyses. a. The hazard ratio was from the multivariable Cox regression model before PSM, with adjustments for all covariates in Table 1. b. The hazard ratio was from a multivariable Cox regression model with the same covariates, with additional adjustment for the propensity score. c. The hazard ratio was from a multivariable Cox regression model with the same covariates with matching according to the propensity score. The analysis included 11,580 patients (5790 who had AF and 5790 who did not). d. The hazard ratio was from the multivariable Cox regression model with the same covariates with inverse probability weighting according to the propensity score