Physicians try to balance the tradeoff between bleeding and stroke risks when prescribing anticoagulants to patients with atrial fibrillation. But from the patient’s perspective, it seems new trumps old and death by bleeding is the worst, according to one study.

Three novel oral anticoagulants (NOACs) are now marketed in the U.S. in addition to warfarin, with a fourth NOAC recently winning a vote of support from the FDA’s advisory panel. While this broadens options for treatments designed to reduce the risk of stroke in patients with atrial fibrillation, it also means physicians and patients must weigh the relative benefits and risks of these drugs in their decision making.

Writing online Nov. 11 in Circulation: Cardiovascular Quality and Outcomes, Mehdi Najafzadeh, PhD, of Brigham and Women’s Hospital in Boston, and colleagues pointed out that physician decisions in this realm are still subjective and the preferences of patients remain unknown. They proposed that incorporating methods to measure patient preference would facilitate decision making and enhance patient-centered care.

They used a Discrete Choice Experiment (DCE) method to assess patients’ attitude about different outcomes from anticoagulant therapy. The system involved an online questionnaire taken by 341 people with a history of cardiovascular disease. The format characterized treatment options based on NOAC trial data, which included the likelihood of nonfatal stroke, nonfatal MI, fatal MI or stroke, minor bleeding, major bleeding, fatal bleeding and the need for blood test monitoring to ensure proper anticoagulation.

Treatment options included old drug, new drug or no drug. The survey also had two questions to check the rationality of the participants’ responses. The rational subset totaled 190 participants.  

Patients were 14 percent less likely to choose an old drug over a new drug, regardless of the actual likelihood of benefit and risk, and they were 28 percent less likely to choose no drug. The rational subset showed no preference for new over old, though.

Patients had the greatest aversion for death by bleeding compared with other outcomes. To avoid a 1 percent increased risk of death by bleeding, they were willing to accept a 2.8 percent increased risk of nonfatal stroke; a 2.2 percent risk of nonfatal MI; a 3.4 percent risk of MI- or stroke-related death; a 16 percent risk of minor bleeding; and a 6 percent increased risk of major bleeding.

“Patients clearly discriminate between different fatal events, suggesting that they have different levels of risk tolerance for those events,” Najafzadeh et al wrote.   

Previously experiencing an MI or stroke also swayed preferences, they found. “Understanding that these patients, on average, are more willing to accept risk in exchange for the benefits of anticoagulants can influence treatment decision in these subpopulations,” they wrote.

They acknowledged that patients might respond differently when faced with real treatment decisions rather than a hypothetical situation.