The objective of the study is to compare the cost-effectiveness of oral anticoagulants among atrial fibrillation patients at an increased stroke risk.

A Markov model was constructed to project the lifetime costs and quality-adjusted survival (QALYs) of oral anticoagulants using a private payer’s perspective. The distribution of stroke risk (CHADS2 score: congestive heart failure, hypertension, advanced age, diabetes mellitus, stroke) and age of the modeled population was derived from a cohort of commercially insured patients with new-onset atrial fibrillation. Probabilities of treatment specific events were derived from published clinical trials. Event and downstream costs were determined from the cost of illness studies. Drug costs were obtained from 2015 National Average Drug Acquisition Cost data.

In the base case analysis, warfarin was the least costly (46241;95%CI,44499–47874)andapixabanhadthehighestQALYs(9.38;95%CI,9.24–9.48QALYs).Apixabanwasfoundtobeacost-effectivestrategyoverwarfarin(incrementalcost-effectivenessratio= 25 816) and dominated other anticoagulants. Probabilistic sensitivity analysis showed that apixaban had at least a 61% chance of being the most cost-effective strategy at willingness to pay value of 100000perQALY.AmongpatientswithCHADS2≥3,dabigatranwasthedominantstrategy.Themodelwassensitivetoefficacyestimatesofapixaban,dabigatran,andedoxabanandthecostofthesedrugs.

Alltheneweroralanticoagulantscomparedweremoreeffectivethanadjusteddosedwarfarin.Ourmodelshowedthatapixabanwasthemosteffectiveanticoagulantinageneralatrialfibrillationpopulationandhasanincrementalcost-effectivenessratio< 50 000/QALY. For those with higher stroke risk (CHADS2≥ 3), dabigatran was the most cost-effective treatment option.