Influence of Sociodemographic Factors and Provider Specialty on Anticoagulation Prescription Fill Patterns and Outcomes in Atrial Fibrillation

• Authors: Wesley T. O'Neal, Emory University; Pratik B. Sandesara, Emory University; J'Neka S. Claxton, Emory University; Richard F. MacLehose, University of Minnesota; Lin Y. Chen, University of Minnesota; Lindsay G. S. Bengtson, Optum; Alanna M. Chamberlain, Mayo Clinic; Faye L. Norby, University of Minnesota; Pamela L. Lutsey, University of Minnesota; Alvaro Alonso, Emory University
• Journal Title: American Journal of Cardiology
• Volume: Volume 122, Number 3
• Publisher: Elsevier: 12 months | 2018-08-01, Pages 388-394
• Type of Work: Article | Post-print: After Peer Review
• Publisher DOI: http://doi.org/10.1016/j.amjcard.2018.04.022
• Subjects: Health Sciences, Medicine and Surgery; Health Sciences, Epidemiology
• Keywords: Science & Technology; Life Sciences & Biomedicine; Cardiac & Cardiovascular Systems; Cardiovascular System & Cardiology; STROKE PREVENTION; TREAT-AF; RISK; PREVALENCE; MANAGEMENT; MORTALITY; SCORE; CARE

Abstract

Early cardiology involvement after atrial fibrillation (AF) diagnosis is associated with increased oral anticoagulant prescription fills and reduced stroke risk. It is unknown if this association varies by race, sex, or education. We examined anticoagulant fills in 223,891 patients with incident nonvalvular AF (mean age = 71 years; 44% women; 84% white; 9% black; 5% Hispanic; 2% Asian) from the Optum Clinformatics database (2009 to 2014). Provider specialty and filled anticoagulant prescriptions 3 months before and 6 months after AF diagnosis were obtained. Poisson regression was used to compute the probability of oral anticoagulant prescription fill and Cox regression was used to estimate the risk of stroke and major bleeding. Cardiology involvement was less likely among nonwhites (white = Referent; black = relative risk = 0.96, 95% confidence interval (0.95 to 0.97); Hispanic = 0.99 (0.98 to 1.00); Asian = 0.95 (0.93 to 0.97)) and women (0.92 (0.91 to 0.93)), but more likely with higher education level (high school or less = Referent; some college = 1.03 (1.02 to 1.04); college or more = 1.08 (1.07 to 1.09)). Patients seen by cardiology providers were more likely to fill anticoagulant prescriptions (Any = 1.67 (1.64 to 1.69); direct oral anticoagulants = 2.59 (2.49 to 2.68); warfarin = 1.38 (1.35 to 1.41)) compared with patients not seen by a cardiology provider. Patients seen by a cardiologist had a reduced stroke risk (hazard ratio = 0.84 (0.79 to 0.88)) and similar bleeding risk (1.01 (0.96 to 1.06)). Outcomes did not vary by race, sex, or education level. In conclusion, although race, sex, and education differences exist in early cardiology involvement after AF diagnosis, the influence of cardiology involvement on anticoagulant prescription fills and AF-related outcomes does not vary by these factors. Initiatives to improve early cardiology referral in nonwhites, women, and those with lower educational attainment may improve AF outcomes.