Emulating a target trial to assess effect modification: an application to obesity in the comparative effectiveness and safety of apixaban versus warfarin in non-valvular atrial fibrillation using electronic health records

Anticoagulation therapy is recommended for patients with non-valvular atrial fibrillation (NVAF) and an increased risk of stroke. Although apixaban showed superiority over warfarin in the Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE) study, little is known about their effects in overweight and obese patients, with concerns that obesity might undermine apixabans effects due to the fixed dosing across body mass index (BMI) groups. We emulated a target trial similar to the ARISTOTLE study using Clinical Practice Research Datalink (CPRD) linked data to estimate the 36-month risk ratios (RR) and risk differences of the effects of apixaban compared to warfarin in NVAF patients in a composite of stroke/systemic embolism (SE), major bleeding and all-cause mortality. In 55,826 patients, apixaban did not differ across groups of BMI in stroke/SE with RR (95% CI) of 1.15 (0.81, 1.62) in normal weight, 1.06 (0.70, 1.61) in overweight and 1.23 (0.69, 2.17) in obese patients. In major bleeding, apixaban was not superior to warfarin in the normal weight group (RR (95% CI) 1.10 (0.76, 1.60) but superior in overweight (RR (95% CI) 0.73 (0.57, 0.93) and obese (RR (95 % CI) 0.67 (0.52, 0.87) groups. In NVAF, the effectiveness and safety of apixaban compared to warfarin were consistent across BMI groups.