2.5 Statistical analysis
The chi-square test and t-test were used to examine the differences in patients’ characteristics between the case group and control group. We performed conditional logistic regression models to estimate the odds ratio (OR) and 95% confidence interval (CI) for UGIB in users of each SSRI, with non-users of that individual SSRI as the reference. The models were adjusted for age, sex, comorbidities, and medications used in order to control the potential confounders. The comorbidities included diabetes, hypertension, dyslipidemia, coronary artery disease, chronic obstructive lung disease, chronic kidney disease, end-stage renal disease, liver cirrhosis, and cancer (Table S1). Previous medications used included antiplatelet agents, oral anticoagulant medications, and nonsteroidal anti-inflammatory drugs (NSAIDs) (Table S2). Regression models were performed repeatedly, using cDDD as the independent variable for each generic SSRI. Same models were also applied using the time between last prescription for SSRIs and the index date as the independent variable, to assess the UGIB risk related to current and recent exposures. All analyses were performed using SAS 9.4 version. Two-sided tests were used to determine the significance at the 5% level.
RESULTS