Anticoagulant therapy is often continued among nursing-home residents with advanced dementia and atrial fibrillation who are nearing the end of life, new research suggests.

“One possible explanation for our findings is that clinicians don’t have clear guidelines about when to continue and when to stop anticoagulants in this population. Our work suggests that we need better evidence about anticoagulation in patients with both atrial fibrillation and dementia to help inform decisions,” Dr. Gregory Ouellet with Yale School of Medicine, in New Haven, Connecticut, said in email to Reuters Health.

Atrial fibrillation (AF) affects almost 20 percent of people with dementia. As the neurodegeneration progresses, the potential benefits of preventing a stroke become increasingly attenuated, the authors note in JAMA Internal Medicine.

To see how often anticoagulant therapy is continued in this patient population, Dr. Ouellet and colleagues used Medicare data to identify 15,217 nursing-home residents with advanced dementia and AF with at least a moderate risk of stroke. The residents died over a four-year period (2014–2017).

A third of these individuals received an anticoagulant in the last six months of life.

Longer time spent in the nursing home and not having Medicaid insurance were more strongly associated with anticoagulant use than stroke risk score, the authors found, and, “counterintuitively,” greater bleeding risk was associated with greater likelihood of anticoagulant use.

“We were surprised to find that patients with several markers of very high short-term mortality, such as difficulty swallowing and weight loss, were more likely to be receiving anticoagulants. This is counterintuitive because the potential benefits are these medications are the lowest in this group,” Dr. Ouellet told Reuters Health.

“These findings underscore the fact that, while practice guidelines contain a well-defined threshold for starting anticoagulation for AF, there is no clear standard for stopping it,” Dr. Ouellet and colleagues add in their article.

“Clinicians are instead asked to engage in shared decision-making with patients and their families. Data about the benefits and harms of therapy are essential to that process. For patients with dementia, little such evidence is available, although the magnitudes of benefits and harms are likely to change substantially as the disease progresses,” they suggest.

“Our work points to the need for high-quality data to inform decision-making about anticoagulation in this population,” they conclude.

The authors of an Editor’s Note say these findings “highlight the lack of a rational strategy for managing anticoagulation in those with limited life expectancy owing to age or illness.”

Dr. Anna Parks and Dr. Kenneth Covinsky of the University of California, San Francisco, say, “Balancing the tradeoffs required for anticoagulation will remain challenging in patients with limited life expectancy. Our goal should be a framework that combines quantitative information with patients’ values to guide clinicians and patients toward individualized and informed decisions.”

The study had no commercial funding. The authors have no relevant disclosures. Dr. Covinsky is associate editor of JAMA Internal Medicine.