Statin therapy and gastrointestinal hemorrhage: a retrospective cohort study with propensity score-matching.

PURPOSE: Whereas some studies suggest that statins exert a gastroprotective effect against gastrointestinal hemorrhage, others report that statin use is associated with increased risk of gastrointestinal hemorrhage. Aim of report: To investigate the risk of gastrointestinal hemorrhage among statin-users compared with non-users. METHODS: This was a retrospective cohort study using clinical, administrative, and pharmacy data encompassing October 2003 to March 2012 from patients enrolled in the San Antonio military health care system. Two treatment groups were defined: statin-users (use for at least 90 days) and non-users (never received statin). A propensity score-matched cohort was generated to match statin-users and non-users based on 82 variables. Main outcome measures were defined by the International Classification of Diseases, ninth revision-clinical modification diagnoses codes or procedural codes for gastrointestinal hemorrhage, gastritis/doudenitis, gastroduodenal ulcers, endoscopy procedures, and endoscopy procedures related to gastrointestinal hemorrhage. RESULTS: A total of 43,438 patients were identified; 13,626 (31.4%) were statin-users and 29,812 were non-users. We propensity score-matched 6342 non-users with 6342 statin-users. The risk of outcomes was similar between the two groups for gastrointestinal hemorrhage (Odds Ratio [OR]: 1.0; 95% confidence interval [95%CI]: 0.91, 1.11); gastrointestinal ulcers (OR: 0.99; 95%CI [0.80, 1.24]); gastritis/duodenitis (OR: 0.92; 95%CI [0.83, 1.02]); and endoscopic procedures (OR: 1.07; 95%CI [0.98, 1.17]). CONCLUSION: Statin use was not significantly associated with either an increased or decreased risk of gastrointestinal hemorrhage. Choice of statin therapy should not be limited in those patients at risk of gastrointestinal hemorrhage.