*2.3. Statistical Analysis*

First, descriptive analyses were performed, and the independent sample T-test (or Kruskal–Wallis test) and chi-square test were used for continuous and categorical variables, respectively. Second, logistic regression analysis was used to examine the relationship between HTN and PAB with SR. The independent variables (IVs) with *p* < 0.2 in the bivariate analysis were selected for the multiple regression models. The multicollinearity among those IVs was controlled by examining their correlation using Spearman's correlation. Because moderate correlations were found between age and occupation (*rho* = 0.34) and between gender and PAB (*rho* = −0.35) (Table S1), the representative variables were selected for multiple analysis models, including age, education, ability to pay for medication, stroke classification, depressive symptoms, and CCI. Third, interaction analysis was used to explore the combined impact of HTN and PAB on SR. Furthermore, the results of the interaction model were visualized via a simple slope analysis using PROCESS Macro of SPSS for moderation analysis. The slope plots were drawn using the estimated values of SR probability for two categories of HTN (yes vs. no) by three levels of PAB (one standard deviation below the mean (−1SD), the mean, and one standard deviation above the mean (+1SD)). All analyses were conducted using SPSS version 22 (IBM Corp., Armonk, NY, USA), and the *p*-value < 0.05 was defined as a statistically significant result.
