|| Checking for direct PDF access through Ovid
Introduction: Spinal cord infarction (SCI) is a rare ischemic event comprising 1% of all strokes. In many cases of SCI, the cause remains undetermined. Case reports have suggested a relationship between cardiac embolism and SCI, but the association between the most common cause of cardiac embolism, atrial fibrillation (AF), and SCI has not been evaluated.Hypothesis: AF is associated with SCI.Methods: We performed a retrospective cohort study using inpatient and outpatient claims data from 2008-2014 on a 5% sample of Medicare beneficiaries. Our predictor variable was AF, ascertained by previously validated ICD-9-CM codes. The primary outcome was SCI, defined as ICD-9-CM diagnosis code 336.1 (vascular myelopathy) among patients who underwent spinal magnetic resonance imaging to rule out a compressive lesion and who did not have a concomitant diagnosis of degenerate joint disease, the most common cause of non-traumatic compressive myelopathy. In sensitivity analyses, we also excluded SCI cases accompanied by codes for traumatic spinal cord injury, spinal cord abscess, or spinal or aortic surgery. Cox proportional hazards analysis was used to assess the relationship between AF and SCI while adjusting for demographic characteristics and vascular risk factors.Results: Among 1,638,461 patients with a mean 3.9 years of follow-up, 423,856 had AF and 22 developed SCI. The annual incidence of SCI was 8.2 (95% confidence interval [CI], 4.4-15.4) per million in patients with AF compared to 2.3 (95% CI, 1.3-4.0) per million per year in those without AF. After adjustment for demographic characteristics and vascular risk factors, AF was associated with a higher risk of SCI (hazard ratio [HR], 4.8; 95% CI, 1.7-13.6). The association between AF and SCI persisted or grew stronger after excluding those with concomitant diagnoses of spinal cord injury, spinal abscess, and spinal or aortic surgery.Conclusions: In Medicare beneficiaries, AF was associated with increased risk of subsequent SCI. These results suggest the need for a thorough evaluation of potential underlying cardioembolic sources in patients with otherwise unexplained SCI.