2.3 Cases and controls
The follow-up duration started on the date of cohort entry, i.e., the date of the first prescription for SSRIs, and ended on the date of hospitalization for UGIB, death, withdrawal from the insurance program, or end of the study period (December 31, 2013), whichever came first. Patients who were admitted for the first time with any of the five codes listed under the non-variceal upper gastrointestinal bleeding category (see Table S1 in the Supplement for International Classification of Diseases, 9th revision, Clinical Modification [ICD-9-CM] codes) comprised the case group[23]. Considering that 16%[24] of patients presenting with melena or hematemesis may experience rebleeding episodes, we excluded patients with previous UGIB. We matched each patient in the case group with three control patients with less than one prescription for SSRIs for age (±1 year), sex, year of cohort entry, and follow-up duration. The index date was defined as the date of hospital admission for UGIB of cases, and the index date of a case was assigned to his or her matched control patient.