We conclude that co-medication in warfarin-treated in-patients is common and should be carefully evaluated to decrease the bleeding risk associated with warfarin therapy.

Abstract:We analysed the occurrence of co-prescribing of potentially interacting drugs during warfarin therapy in the community-dwelling population of Finland. We identified drugs having interaction potential with warfarin using the Swedish Finnish INteraction X-referencing drug-drug interaction database (SFINX) and obtained data on drug purchases from the nationwide Prescription Register. We defined warfarin users as persons purchasing warfarin in 2010 (n = 148,536) and followed them from their first prescription in 2010 until the end of the calendar year. Co-prescribing was defined as at least 1-day overlap between warfarin and interacting drug episodes. In addition, we identified persons who initiated warfarin therapy between 1 January 2007 and 30 September 2010 (n = 110,299) and followed these incident users for a 3-month period since warfarin initiation. Overall, 74.4% of warfarin users were co-prescribed interacting drugs. Co-prescribing covered 46.4% of the total person-years of warfarin exposure. Interacting drugs that should be avoided with warfarin were co-prescribed for 13.4% of warfarin users. The majority of the co-prescriptions were for drugs that are not contraindicated during warfarin therapy but require special consideration. Among incident users, 57.1% purchased potentially interacting drugs during the 3-month period after initiation, while 9.0% purchased interacting drugs that should be avoided with warfarin. To conclude, the occurrence of co-prescribing of potentially interacting drugs was high during warfarin therapy. Our findings highlight the importance of close monitoring of warfarin therapy and the need for further studies on the clinical consequences of co-prescribing of interacting drugs with warfarin.

Clinical significance of potential interaction between warfarin and statins is unclear. Our objective was to determine whether use of statins as a class or use of simvastatin modulates the rate of bleeding requiring hospitalization among new warfarin users. Using Finnish healthcare databases, we identified a cohort of 101,588 warfarin initiators between 1 January 2009 and 30 June 2012. By the end of 2012, these patients accumulated 92,695 person-years of exposure to warfarin-only and 60,253 years of exposure to warfarin-with-statin. The outcome was a composite of gastrointestinal, intracranial or other bleeding leading to hospitalization. A Poisson generalized estimating equation model was employed to estimate rate ratios (RR) and their 95% confidence intervals (CI) for exposure to warfarin-with-statin compared to warfarin-only and to allow multiple episodes per patient and time-dependent covariates. In multivariable models, we found no difference in the bleeding rate in association with exposure to any statin (multivariable-adjusted RR = 0.98, 95% CI 0.89-1.07) or to simvastatin (RR = 1.01, 95% CI 0.91-1.11) with warfarin compared to exposure to warfarin-only. We conclude that concomitant use of statins and warfarin was not associated with an increased rate of bleeding requiring hospitalization.