Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

PPO 13-387 – HSR Study

PPO 13-387
Cost-Effectiveness of Dabigatran and Warfarin for Veterans with AFib: Pilot Study
Julia C Prentice, PhD MSPH
VA Boston Healthcare System Jamaica Plain Campus, Jamaica Plain, MA
Boston, MA
Funding Period: April 2014 - March 2015
An estimated 2.7 million individuals in the United States were diagnosed with atrial fibrillation (AFib) in 2010. These individuals are estimated to have a five-fold increased risk of ischemic stroke. A key component of AFib treatment is the prevention of blood clots to reduce the risk of stroke. Vitamin K antagonists (e.g. warfarin) have successfully reduced the risk of stroke by two-thirds but the effectiveness of these methods are hampered by narrow therapeutic ranges and the need for monthly laboratory monitoring for dose adjustment to prevent over-anticoagulation and subsequent hemorrhage. Since 2010, two new classes of oral NoACs thrombin inhibitors (e.g. dabigatran) and Xa inihibitors entered the market. An advantage of these drugs is that they do not require regular laboratory monitoring for dose adjustment. Clinical trials provide good evidence that the NoACs are either superior or not inferior to warfarin on a variety of outcomes including stroke, mortality and hemorrhage risk. A disadvantage of NoACs is cost but initial cost-effectiveness studies based on clinical trial results have concluded the NoACs may be cost-effective. It is unknown whether the health and cost advantages of NoACs found in clinical trials will be maintained in standard clinical practice since these trials enroll healthier patients and maintain better adherence. Our future proposal will be a comparative and cost-effectiveness study of warfarin and dabigatran in veterans diagnosed with AFib.

The first NoAC on the VA national formulary, dabigatran, was approved starting in FY2012. Rivaroxaban and Apixaban have since been included. The first objective examines whether NoAC utilization is common enough for adequate power in the future proposal and the second objective examines NoAC practice pattern variation.

Objective 1. In one test year of data, identify a cohort of VA patients diagnosed with AF and prescribed warfarin or dabigatran. Determine number of years of data needed to achieve adequate expected statistical power in future proposal.

Objective 2. In one test year of data, demonstrate geographic practice pattern variation in likelihood of receiving dabigatran as a function of medical center effects and distance to the nearest medical center. Control for known factors affecting choice of anticoagulation therapy.

This pilot was a retrospective observational study of secondary data from patient-level administrative and claims data from VA. VA pharmacy data was used to count how many patients were prescribed warfarin or a NoAC in FY 2012-2013. We estimated sample sizes for the future study based on this cohort and calculated expected statistical power based on observed stroke, hemorrhage and mortality rates. For objective 2, we used a logistic regression model to predict the probability of receiving warfarin compared to a NoAC as a function of demographics, comorbidities, distance to the nearest medical centers and facility fixed effects.

Between FY12 and FY14, there were 47,785 individuals who were diagnosed with Afib and then started either warfarin or a NoAC. In 2012, 86% started on warfarin but in 2014 only 74% did. Twelve to 13% started on dabigatran in each year. The proportion starting on rivaroxaban increased from 2% in FY 12 to 9% in FY 14.

There is significant prescribing variation between providers as NoACs entered the national VA formulary. A provider warfarin prescribing rate was calculated by dividing the number of warfarin prescriptions by the total number of warfarin and NoAC prescriptions in each month. The mean warfarin prescribing rate decrease from 96% at the beginning of FY 12 to 71% at the end of FY 14. There is also a steady increase in the variance of this variable over time so the variation in provider prescribing decisions is increasing over time.

Results from models that use this provider-level prescribing rate to predict the likelihood of an individual initiating warfarin compared to a NoAC suggest the proposed instrument for the proposed comparative and cost-effectiveness study has plenty of power. The F-statistic in these models is 1210 which is well above the Staiger/Stock criteria of F>10. Younger individuals, those with elevated creatinine clearance impairment and individuals with a higher CHADS score are significantly more likely to initiate warfarin. Veterans who visit facilities that are in the top quartile of %TTR were also significantly more likely to initiate warfarin compared to individuals who visit facilities with less TTR control.

The emergence of novel anticoagulants (NoAC) has provided clinicians with alternatives to Vitamin K antagonists for anticoagulation therapy. The completed pilot project provided essential data on the likelihood that the larger comparative and cost-effectiveness study comparing warfarin to NoACs will succeed. This larger study will increase access to the most effective treatments for veterans and help VA policymakers develop evidence-based guidelines for AFib treatment.

External Links for this Project

NIH Reporter

Grant Number: I21HX001475-01

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


None at this time.

DRA: Cardiovascular Disease, Aging, Older Veterans' Health and Care
DRE: Treatment - Comparative Effectiveness
Keywords: none
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.