Underuse of oral anticoagulation for individuals with atrial fibrillation in a nursing home setting in France: comparisons of resident characteristics and physician attitude

Objectives: To describe the characteristics of nursing home residents diagnosed with atrial fibrillation (AF) and eligible for oral anticoagulants who did not receive these drugs and to detail the conditions that physicians who decide not to prescribe anticoagulants take into account.

Design: Cross-sectional.

Setting: Nursing home.

Participants: Nursing home residents with a history of AF (N = 1,085).

Measurements: Data were collected on clinical characteristics, geriatric syndromes, and antithrombotic regimen. Multivariate logistic regression was used to identify factors associated with nonprescription of anticoagulants. A standardized questionnaire was submitted to physicians in charge of patients with AF, to detail conditions associated with their medical decision not to prescribe anticoagulants.

Results: History of AF was present in 1,085 nursing home residents (10.1%), mean age 87, with a mean CHA2DS2-VASc score of 5.1 ± 1.4. Of these residents with AF, 544 (50.1%) did not receive anticoagulants. Recurrent falls (odds ratio (OR) = 4.9, 95% confidence interval (CI) = 2.4-9.9, P < .001), past history of bleeding (OR = 3.62, 95% CI = 1.54-8.51, P = .003), paroxysmal AF (OR = 3.5, 95% CI = 1.83-6.66, P < .001), and advanced age (OR = 1.1, 95% CI = 1.01-1.17, P = .02) were significantly associated with not prescribing anticoagulants. Recurrent falls (47%), cognitive impairment (22.6%), and advanced age (16.4%) were the main reasons for not prescribing anticoagulants.

Conclusion: The prevalence of AF in a cohort of very old nursing home residents was 10%. Anticoagulation was prescribed in fewer than 50% of eligible cases despite high individual risk of stroke. Geriatric syndromes, especially falls and cognitive disorders, were the main reported contraindications for prescribing anticoagulants. Physicians caring for those residents wrongly thought that paroxysmal AF caused fewer thromboembolic events than permanent AF, which explains lower rates of anticoagulant prescription in individuals with paroxysmal AF.