• Retrospective study using data from VA linked with Medicare data from CMS from January 1, 2013 to
    December 31, 2017
  • Patients were assigned to 1 of the 3 groups*:
    • Non-obese: BMI <30 kg/m2 or body weight <100 kg
    • Obese (non-morbid): BMI of 30 to 40 kg/m2 or body weight of 100 to 120 kg
    • Obese (morbid): >40 kg/m2 or body weight >120 kg
  • Stabilized IPTW was used to balance patient characteristics among the apixaban and warfarin cohorts
  • Incidence rates per 100 person-years were calculated for stroke/SE and MB outcomes
  • Cox proportional hazards models were used to estimate the risks of stroke/SE and MB
  • Interaction analysis evaluated the treatment effect across non-obese, obese (non-morbid), and obese (morbid)
    patients
  • Patients were followed from the index date until the earliest of the following: 30 days after the date of discontinuation, switch, death, end of continuous medical enrollment, or end of study period