The impact of interacting drugs on dispensed doses of warfarin in the Swedish population: A novel use of population based drug registers.

To investigate the impact of interacting drugs on the dispensed doses of warfarin in the Swedish population. This was a retrospective, cross-sectional population based register study of patients being dispensed warfarin. Warfarin doses were estimated in different age groups, in men and women, and in patients using interacting drugs. The influence of interacting drugs on the dispensed warfarin dose was analyzed using multiple regression. All 143,729 patients dispensed warfarin were analyzed. The dispensed dose of warfarin was highest in patients 30-39 years old and decreased with age. Co-medication with carbamazepine, simvastatin, paracetamol, amiodarone, fluconazole, lactulose, or bezafibrate was associated with significant changes in dispensed warfarin doses, by +40%, -3.4%, -7.3%, -8.2%, -8.8%, -9.0%, and -9.7%, respectively. After adjustment for age and gender, sulfamethoxazole was also found to significantly alter the dispensed warfarin dose (-6.1%). We provide new support for the previous scarce evidence of interactions between warfarin and carbamazepine, bezafibrate, and lactulose. Initiation or discontinuation of bezafibrate or lactulose in a patient on warfarin should warrant close clinical monitoring. The marked increased warfarin requirement associated with carbamazepine use supports moving from a more conservative reactive towards a proactive strategy including preventive warfarin dose adjustments to avoid potential adverse effects.