Describe and compare healthcare resource utilization (HCRU) among non-valvular atrial fibrillation (NVAF) patients newly treated with apixaban vs. VKA (vitaminK antagonist), rivaroxaban or dabigatran.

All adult patients with NVAF newly initiating apixaban, VKA, rivaroxaban or dabigatran between January 2014 and December 2016 (absence of OAC use in the previous 24months) were identified using the SNIIRAM database. Different types of HCRU were described: medical visits, nurse visits, concomitant drugs, biology dosings, medical procedures and hospital stays. HCRU were compared between groups with generalized linear model adjusted on propensity score, taking account of the follow-up time of patients as offset.

Final cohorts included 87,565 apixaban, 112,628 VKA, 100,063 rivaroxaban, and 21,245dabigatran patients. Over the follow-up period, apixaban patients had an average of 1.8 medical visits per month (SD: 1.9), mostly with general practitioners (89.9%) and cardiologists (52.2%). Mean number of nurse visits was 4.5 per month (SD: 8.5) for 66.3% of patients treated with apixaban. Mean number of concomitant drugs packages was 13.4 per month (SD: 18.0), and most frequent drug classes were antiarrhythmics, beta-blocking agents, analgesics in 68.2%, 64.1%, 60.5% respectively. Half of patients treated with apixaban had monthly dosing of creatinine, glucose and transaminases (56.5%, 55.5%, and 53.1%). Most frequent medical procedures were Doppler, thorax radiography and electrocardiography, 48.5%, 27.6%, and 26.8% and 47.7% of patients had at least one hospitalization. For all HCRU types, apixaban patients had significantly fewer HCRU than VKA patients. For drug claims, biology, and medical procedures, apixaban patients had significantly fewer HCRU than rivaroxaban and dabigatran patients.

This large real-world study found lower adjusted HCRU for apixaban compared to VKA and, to a lesser extent, compared to rivaroxaban and dabigatran.