Stroke: Dabigatran cost-effective in AF patients with prior stroke, TIA
Warfarin, the war horse for treating this patient population, is inexpensive but requires monitoring. The alternative dabigatran (Pradaxa, Boehringer Ingelheim) is more expensive but less limited than warfarin. Two studies published in 2011 compared the cost effectiveness of these two drugs but came to different conclusions, with one study estimating the incremental cost-effectiveness ratio (ICER) of dabigatran at $12,000 per quality-adjusted life years (QALYs) and another pegging it at more than $50,000 (Ann Intern Med 2011;154:570–571 and Circulation 2011;123:2562–2570, respectively).
Hooman Kamel, MD, of the department of neurology and neurosciences at Weill Cornell Medical College in New York City, and colleagues designed their study using a Markov decision model and sensitivity analyses in an attempt to add clarity for physicians who prescribe these drugs for secondary stroke prevention. “We used broadly similar inputs [as the previous two studies] to ensure comparability of our results but made several assumptions that more specifically reflect the natural history of patients with prior stroke and TIA,” the authors wrote.
For the analysis, they used data on the risks of adverse events from the RE-LY trial and warfarin trials for AF patients. Their base case was a hypothetical cohort of patients, 70 years old or older, with nonvalvular AF, either a prior stroke or TIA, and no contraindications to anticoagulation. They adjusted mortality rates for age, presence of prior stroke or TIA and type of therapy, and determined quality-adjusted life expectancy by multiplying the probability of adverse events by their expected utilities. Because dabigatran lacked direct utility data, they assigned it the same utility as ximelagatran (Exanta, AstraZeneca), in keeping with other studies.
For the cost component, the researchers selected the published cost of dabigatran, with total cost based on wholesale price and standard office visits. Warfarin therapy cost covered Medicare reimbursement for 90 days of anticoagulation management and 14 international normalized ratio tests. The cost of hospitalization and ongoing costs from adverse events were taken from federal data. All costs were in 2010 dollars.
They compared adjusted-dose warfarin with an international normalized ratio (INR) target of 2 to 3 with 150 mg dabigatran taken twice daily. Base case results favored dabigatran, showing a quality-adjusted life expectancy (QALY) of 4.27 QALYs for dabigatran compared with 3.91 QALYs for warfarin. Their analysis concluded that dabigatran provided 0.36 additional QALYs at a cost of $9,000 for an ICER of $25,000.
But sensitivity analyses showed the cost-effectiveness of dabigatran was inversely related to the quality of INR control achieved with warfarin therapy. In a data supplement, the authors noted that in centers with a time in the therapeutic INR range of more than 73 percent, dabigatran was not cost-effective at a threshold of $50,000. They described this as “uncommonly good INR control.”
The Monte Carlo analysis showed that dabigatran was cost-effective in 57 percent of simulations using a threshold of $50,000 per QALY and 78 percent of simulations using a threshold of $100,000 per QALY. It was not cost-effective compared with warfarin if the relative risk of stroke was more than 0.92.
“In an analysis using data from a randomized trial, we have found that dabigatran is a cost-effective alternative to warfarin in patients with AF and prior stroke or TIA,” Kamel and colleagues wrote. “The cost-effectiveness of dabigatran appears to depend on the adequacy of warfarin management with a greater advantage for dabigatran in centers with poor international normalized ratio control in warfarin-treated patients.”
In the data supplement, they added that dabigatran may be preferable in cases where patient access to an anticoagulation center is challenging or where the care provider is less experienced at managing warfarin therapy.
The authors noted that their study relied on data from a single trial, RE-LY, and recommended future studies based on new data as they emerge.