Bleeding complications in patients on celecoxib and warfarin1

Authors


  • 1

    The authors have no financial disclosures related to this manuscript. The results of this study were presented at the American College of Rheumatology meeting in October 2004 as a poster presentation.

Lorinda Chung, 1000 Welch road, Suite no. 203, Stanford, CA 94305, USA. Tel.: (650)723 6961; fax: (650)725 8418; e-mail: shauwei@stanford.edu

Summary

Purpose:  Non-selective non-steroidal anti-inflammatory drugs (nNSAIDs) used in combination with warfarin are associated with an approximately 3-fold increased risk of upper gastrointestinal bleeding (UGIB) compared with warfarin alone. Celecoxib, a selective inhibitor of cyclo-oxygenase 2 (COX-2), is associated with less gastric mucosal injury and platelet dysregulation than nNSAIDs. We compared rates of bleeding complications in patients taking celecoxib and warfarin with those taking warfarin alone.

Subjects and Methods:  We performed a retrospective analysis using data from our Protime Clinic and pharmacy databases from January 2001 to April 2004. We identified 123 patients who took celecoxib and warfarin concurrently (overlap group). We compared rates of bleeding complications in this group with 1022 control patients who were taking warfarin alone. Bleeding complications were defined as major if they resulted in hospitalization, blood transfusion or death.

Results:  During approximately 1063 months of exposure to both celecoxib and warfarin, 10 bleeding complications were identified, only one of which was considered major. No patients had UGIB. In the control group, 116 bleeding complications were identified over approximately 16 520 months of exposure to warfarin alone, with 101 minor and 15 major events, including six episodes of UGIB. The relative risk of all bleeding complications was 1·34 (95% CI: 0·70–2·57) in the overlap vs. control groups, and for major bleeds was 1·04 (95% CI: 0·14–7·85).

Conclusions:  There is a mild but non-significant increase in bleeding complications in patients taking celecoxib and warfarin compared with those taking warfarin alone.

Ancillary