Open Access Highly Accessed Research article

Characteristics affecting oral anticoagulant therapy choice among patients with non-valvular atrial fibrillation: a retrospective claims analysis

Azza AbuDagga1, Judith J Stephenson1, An-Chen Fu1*, Winghan Jacqueline Kwong2, Hiangkiat Tan1 and William S Weintraub3

Author Affiliations

1 HealthCore, Inc., 800 Delaware Ave, 5th floor, Wilmington, DE 19801, USA

2 Daiichi Sankyo, Inc., 2 Hilton Court, Parsippany, NJ 07054, USA

3 Christiana Care Health System, 4755 Ogletown-Stanton Road, Suite 2E99, Newark, DE 19718, USA

For all author emails, please log on.

BMC Health Services Research 2014, 14:310  doi:10.1186/1472-6963-14-310

Published: 17 July 2014



Dabigatran is one of the three newer oral anticoagulants (OACs) recently approved in the United States for stroke prevention in non-valvular atrial fibrillation (NVAF) patients. The objective of this study was to identify patient, healthcare provider, and health plan factors associated with dabigatran versus warfarin use among NVAF patients.


Administrative claims data from patients with ≥2 NVAF medical claims in the HealthCore Integrated Research Database between 10/1/2009 and 10/31/2011 were analyzed. During the study intake period (10/1/2010 - 10/31/2011), dabigatran patients had ≥2 dabigatran prescriptions, warfarin patients had ≥2 warfarin and no dabigatran prescriptions, and the first oral anticoagulant (OAC) prescription date was the index date. Continuous enrollment for 12 months preceding (“pre-index”) and ≥ 6 months following the index date was required. Patients without pre-index warfarin use were assigned to the ‘OAC-naïve’ subgroup. Separate analyses were performed for ‘all-patient’ and ‘OAC-naïve’ cohorts. Multivariable logistic regression (LR) identified factors associated with dabigatran versus warfarin use.


Of 20,320 patients (3,019 dabigatran and 17,301 warfarin) who met study criteria, 27% of dabigatran and 13% of warfarin patients were OAC-naïve. Among all-patients, dabigatran patients were younger (mean 67 versus 73 years, p < 0.001), predominantly male (71% versus 61%, p < 0.001), and more frequently had a cardiologist prescriber (51% versus 30%, p < 0.001) than warfarin patients. Warfarin patients had higher pre-index Elixhauser Comorbidity Index (mean: 4.3 versus 4.0, p < 0.001) and higher ATRIA bleeding risk score (mean: 3.0 versus 2.3, p < 0.001). LR results were generally consistent between all- and OAC-naïve patients. Among OAC-naïve patients, strongest factors associated with dabigatran use were prescriber specialty (OR = 3.59, 95% CI 2.68-4.81 for cardiologist; OR = 2.22, 95% CI 1.65-2.97 for other specialist), health plan type (OR = 1.47 95% CI 1.10-1.96 for preferred provider organization), and prior ischemic stroke (OR = 1.42, 95% CI 1.06-1.90). Older age decreased the probability of dabigatran use.


Beside patient characteristics, cardiology specialty of the prescribing physician and health plan type were the strongest factors associated with dabigatran use.

Non-valvular atrial fibrillation; Dabigatran; Warfarin; Oral anticoagulation; Healthcare