Containing supporting information.

<p>Table A1. ARISTOTLE inclusion and exclusion criteria applied to CPRD Aurum. Table A2. Efficacy outcomes results from ARISTOTLE. Table A3. Effectiveness outcomes results in the CPRD Aurum ARISTOTLE-analogous cohort. Table A4. Bleeding outcomes and net clinical outcomes results from ARISTOTLE RCT. Table A5. Bleeding outcomes and net clinical outcomes results in the CPRD Aurum ARISTOTLE-analogous cohort. Table A6. Effectiveness outcomes results in the CPRD Aurum ARISTOTLE-analogous cohort using the on-treatment censoring scheme. Table A7. Treatment status of apixaban and warfarin users in CPRD Aurum ARISTOTLE-analogous cohort during 2.5 years of follow-up. Table A8. Characteristics of apixaban and warfarin users in CPRD Aurum ARISTOTLE-analogous cohort by treatment persistence during 2.5 years of follow-up. Table A9. Effectiveness outcomes results in the CPRD Aurum ARISTOTLE-analogous cohort using later study start date (1 January 2014). Table A10. Summary of noninterventional studies comparing apixaban and warfarin in AF patients.</p> <p>(DOCX)</p>