Bleeding Risk, Physical Functioning, and Non-use of Anticoagulation Among Patients with Stroke and Atrial Fibrillation

Abstract

Background: Atrial fibrillation (AF) is common among people with stroke. Anticoagulation medications can be used to manage the deleterious impact of AF after stroke, however may not be prescribed due to concerns about post-stroke falls and decreased functioning. Thus, the purpose of this study was to identify, among people with stroke and AF, predictors of anticoagulation prescription at hospital discharge. Methods: This is a secondary analysis of a retrospective cohort study of data retrieved via medical records, including: National Institutes of Health Stroke Scale score; Functional Independence Measure (FIM) motor score (motor or physical function); ambulation on 2nd day of hospitalization; Morse Falls Scale (fall risk); and HAS-BLED score (Hypertension; Abnormal renal and liver function; Stroke; Bleeding; Labile INRs; Elderly > 65; and Drugs or alcohol). Data analyses included bivariate comparisons between people with and without anticoagulation at discharge. Logistic-regression modeling was used to assess predictors of discharge anti-coagulation. Results: There were 334 subjects included in the analyses, average age was 75 years old. Anticoagulation was prescribed at discharge for 235 (70%) of patients. In the adjusted regression analyses, only the FIM motor score (adjusted OR = 1.015, 95%CI 1.001-1.028) and the HAS-BLED score (adjusted OR = 0.36, 95%CI 0.22-0.58) were significantly associated with anticoagulation prescription at discharge. Conclusion: It appears that in this sample, post-stroke anti-coagulation decisions appear to be made based on clinical factors associated with bleed risk and motor deficits or physical functioning. However, opportunities may exist for improving clinician documentation of specific reasoning for non-anticoagulation prescription.