Abstract 38: Comparison of Gastrointestinal Bleeding Risk Among Different Statin Exposures With Warfarin: An Electronic Health Record-based Retrospective Cohort Study
Introduction: Patients who require both warfarin and statin are frequently seen in vascular clinics. Gastrointestinal bleeding (GIB) is one of the most frequently occurring complications when a patient is administered warfarin, and this complication hinders continuous administration of warfarin.
Hypothesis: GIB risk may differ among different statin exposures with warfarin.
Methods: We observed patients who were exposed to one of four statins (pravastatin, simvastatin, atorvastatin, or rosuvastatin) and warfarin concomitantly for up to 2 years (730 days). Observation was ended when GIB occurred or observation was censored. To evaluate the risk of co-administration of different statins and warfarin on GIB, within-class comparison was used. We performed 1:1 matching using a propensity score for comparison of each single statin and others. Kaplan-Meier analysis with log-rank test and Cox proportional hazard regression were conducted for comparison.
Results: We identified 1,686 patients who were administered statin concomitantly with warfarin. The patients were classified by each statin: 287 patients were exposed to pravastatin, 317 to simvastatin, 716 to atorvastatin, and 366 to rosuvastatin. Mean levels of total cholesterol during the observation period were 167 ± 44, 155 ± 36, 145 ± 35, and 139 ± 37 mg/dl, respectively (p<0.001). On Kaplan-Meier curve analyses, GIB was less frequent in the pravastatin group than in others (p<0.05). It was more frequent in the rosuvastatin group than in others (p=0.009). In the Cox proportional hazard regression analysis, age, sex, Charlson comorbidity index, nonsteroidal anti-inflammatory drug exposure, steroid exposure, antiplatelet exposure, and the proportion of time in the therapeutic range of warfarin international normalized ratio were adjusted. Hazard ratio (HR) of GIB on statin exposure in the rosuvastatin group was significantly high (HR, 5.0 [95% confidence interval, 1. 1-23.3], p=0.038) while the ratio in the pravastatin group tended to low (0.3 [0.1-1.0], p=0.051).
Conclusions: Rosuvastatin administration has a relatively high risk of GIB among patients who are concomitantly administered warfarin, whereas pravastatin administration appears to carry a low risk.
Author Disclosures: J. Lee: None.
- © 2016 by American Heart Association, Inc.