Comparison of all-cause costs and healthcare resource use among patients with newly-diagnosed non-valvular atrial fibrillation newly treated with oral anticoagulants

Compare costs and healthcare resource utilization (HCRU) among newly-diagnosed non-valvular atrial fibrillation (NVAF) patients newly treated with dabigatran vs apixaban, rivaroxaban, or warfarin.Newly-diagnosed adult NVAF patients initiating dabigatran, apixaban, rivaroxaban, or warfarin (index event) between October 1, 2010-December 31, 2014 were identified using MarketScan claims data, and followed until medication discontinuation, switch, inpatient death, enrollment end, or study end (December 31, 2015). Dabigatran patients were propensity-score matched 1:1 separately with apixaban, rivaroxaban, and warfarin patients. Per-patient-per-month (PPPM) all-cause cost, HCRU, and 30-day re-admissions were reported. Costs were analyzed using generalized linear models.Final cohorts, each matched with dabigatran patients, included 8,857 apixaban patients, 26,592 rivaroxaban patients, and 33,046 warfarin patients. Dabigatran patients had lower adjusted PPPM total healthcare, inpatient, and outpatient costs compared to rivaroxaban ($4,093 vs $4,636, $1,476 vs $1,862, and $2,016 vs $2,121, respectively, all p ≤ .001) and warfarin ($4,199 vs $4,872, $1,505 vs $1,851, and $2,049 vs $2,514, respectively, all p .001). Adjusted costs were similar for dabigatran and apixaban. Dabigatran patients had significantly fewer hospitalizations, outpatient visits, and pharmacy claims than rivaroxaban patients (0.06 vs 0.07, 4.84 vs 4.96 and 4.80 vs 4.93, respectively, all p .020) and warfarin patients (0.06 vs 0.07, 4.77 vs 6.88, and 4.76 vs 5.89, respectively, all p .001). Dabigatran patients had similar hospitalizations to apixaban, but higher outpatient visits (4.70 vs 4.31) and pharmacy claims (4.86 vs 4.61), both p .001.This real-world study found adjusted all-cause costs were lower for dabigatran compared to rivaroxaban and warfarin patients and similar to apixaban patients.