By Annette M. Boyle
PALO ALTO, CA – Cardiologists may overtreat veterans newly diagnosed with atrial fibrillation and atrial flutter, while primary care physicians may undertreat them, according to research recently published in the American Heart Journal.1
“Overall, the differences are pretty striking,” lead author Mintu Turakhia, MD, told U.S. Medicine. “After adjustments, we found that patients seen by cardiologists were twice as likely to receive warfarin.”
At the low end of the CHADS2 score, used to predict the risk of stroke in veterans with nonrheumatic atrial fibrillation (AF), researchers found that cardiologists prescribed warfarin, while primary care physicians did not. “A patient with a CHADS2 stroke of 2 or higher should be anticoagulated,” said Turakhia. “Maybe even a patient with a 1 should be on anticoagulants. The majority of cardiologists in our study anticoagulated patients even with CHADS2 scores of 0.”
Overtreating vs. Undertreating
“Cardiologists may be overtreating, but primary care physicians seem to be undertreating,” Turakhia noted. PCPs did not prescribe anticoagulants to the majority of new AF patients until their CHADS2 scores reach 4 or higher.
“Primary care physicians used anticoagulants more as CHADS2 scores rose. So we saw that as the risk increased the gap in treatment went down,” he added.
Even so, marked differences in treatment patterns remained, with cardiologists prescribing warfarin to more than two-thirds of their patients, regardless of CHADS2 score. PCPs, by contrast, did not prescribe warfarin to more than 58% of patients, even those who had the highest CHADS2 scores of 5 and 6.
Researchers used the Veterans Health Administration health records and claims data to identify patients newly diagnosed with AF between October 2004 and November 2008 and had one or more primary care or cardiology appointments within 90 days of diagnosis. Of the 141,642 patients who met the criteria, the mean age was 72.3, 98.55 of whom were male. Slightly more than one-quarter received outpatient care from a cardiologist.
During the study period, the use of warfarin in patients treated only by primary care physicians declined from an average of 51.6% to 44%. At the same time, the use of warfarin in patients receiving care from a cardiologist remained steady.
While patients seen by cardiologists had more co-morbidities and higher mean CHADS2 scores, the findings persisted across a series of adjusted models and subgroup analyses. There was a higher unadjusted and adjusted warfarin use in patients treated by cardiologists, regardless of bleeding risk, receipt of rate control drugs or prescription of antiarrhythmic medications. In addition to adjusting for other clinical factors, the researchers accounted for variance in VA benefits and distance to clinics.
“There is a big practice gap between the two groups,” Turakhia said. “There’s been a movement in cardiology away from aspirin to using anticoagulants in lower risk patients as we have seen major improvement in preventing bleeding with warfarin and increased use of new anticoagulants.”
Much of the movement toward using anticoagulants in patients with lower CHADS2 scores has been driven by new recommendations from professional societies, particularly those that have embraced the novel anticoagulants.
“The new guidelines from the American College of Chest Physicians have an ‘all or none’ approach,” Turakhia pointed out. “If a patient has a CHADS2 score of 0, they don’t need aspirin. If it’s higher, they need warfarin or another anticoagulant. Using the European Society of Cardiology recommendations, an even larger proportion would be recommended for anticoagulation.”
The VA, however, has taken a more measured approach to the use of new anticoagulants. “We’re not systematically switching tens of thousands of patients to new drugs. The VA has been careful to match patients to the groups studied in the trials,” he noted.
Differing Views of Risk
Different perceptions of risks of thromboembolic events and bleeding also might explain the greater use of anticoagulants by cardiologists. Anticoagulation of patients with CHADS2 scores of 0 might indicate that patients at low risk of stroke are being exposed to greater risk of other adverse outcomes by treatment. On the other hand, the researchers note, “the higher observed rates of anticoagulation in the low-risk groups may merely reflect specialists’ perception of the limitations of the CHADS2 score in guiding warfarin therapy.”
Primary care physicians may evaluate use of anticoagulants on different bases. “The primary care physician may have a higher threshold to anticoagulate after consideration of potential harms of anticoagulation, given their emphasis on the general care of their patient,” stated the authors. Non-cardiologists also might prefer to observe patients for a longer period before prescribing anticoagulants.
The researchers noted that the reason for clinical evaluation also might affect the prescription of anticoagulants. While all the patients in the study saw a PCP or cardiologist within three months of diagnosis with AF, there was no way to tell whether the patient sought an appointment because of the diagnosis or for other reasons. If a patient sees a primary care physician for some reason other than AF, management of AF might not be discussed during the visit “especially in the setting of busy practices with competing demands” that characterize primary care, Turakhia suggested.
“It’s important to understand these differences in treatment. We need to ensure that all veterans have access to quality care, regardless of site of treatment or specialty of treating physician,” he said.
1 Turakhia MP, Hoang DD, Xu X, Frayne S, Schmitt S, Yang F, et al. “Differences and trends in stroke prevention anticoagulation in primary care vs cardiology specialty management of new atrial fibrillation: The Retrospective Evaluation and Assessment of Therapies in AF (TREAT-AF) study.” American Heart Journal. 2013;165(1):93-101