Development and Validation of a Prediction Model for Atrial Fibrillation Using Electronic Health Records
This article requires a subscription or purchase to view the full text. If you are a subscriber or member, click the login link or the subscribe link in the top menu above to access this article.
Author + information
- Received June 11, 2019
- Accepted July 22, 2019
- Published online November 18, 2019.
Article Versions
- previous version (October 2, 2019 - 11:00).
- You are viewing the most recent version of this article.
Author Information
- Olivia L. Hulme, MDa,∗,
- Shaan Khurshid, MDb,∗,
- Lu-Chen Weng, PhDa,
- Christopher D. Anderson, MD, MMScc,d,
- Elizabeth Y. Wang, MDa,
- Jeffrey M. Ashburner, PhD, MPHe,f,
- Darae Ko, MD, MScg,
- David D. McManus, MD, MSch,
- Emelia J. Benjamin, MD, ScMi,j,
- Patrick T. Ellinor, MD, PhDa,k,
- Ludovic Trinquart, PhDi,l and
- Steven A. Lubitz, MD, MPHa,k,∗ (slubitz{at}mgh.harvard.edu)
- aCardiovascular Research Center, Massachusetts General Hospital, Boston, Massachusetts
- bCardiology Division, Massachusetts General Hospital, Boston, Massachusetts
- cCenter for Genomic Medicine, Massachusetts General Hospital, Boston, Massachusetts
- dJ.P. Kistler Stroke Research Center, Massachusetts General Hospital, Boston, Massachusetts
- eDivision of General Internal Medicine, Massachusetts General Hospital, Boston, Massachusetts
- fDepartment of Medicine, Harvard Medical School, Boston, Massachusetts
- gDepartment of Medicine, Boston University Medical Center, Boston, Massachusetts
- hDepartments of Medicine and Quantitative Health Sciences, University of Massachusetts Medical School, Worcester, Massachusetts
- iBoston University and National Heart, Lung, and Blood Institute Framingham Heart Study, Framingham, Massachusetts
- jSections of Preventive Medicine and Cardiovascular Medicine, Department of Medicine, Boston University School of Medicine, and Department of Epidemiology, Boston University School of Public Health, Boston, Massachusetts
- kCardiac Arrhythmia Service, Massachusetts General Hospital, Boston, Massachusetts
- lDepartment of Biostatistics, Boston University School of Public Health, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Steven A. Lubitz, Cardiac Arrhythmia Service and Cardiovascular Research Center, Massachusetts General Hospital, 55 Fruit Street, GRB 109, Boston, Massachusetts 02114.
Central Illustration
Abstract
Objectives This study sought to determine whether the risk of atrial fibrillation AF can be estimated accurately by using routinely ascertained features in the electronic health record (EHR) and whether AF risk is associated with stroke.
Background Early diagnosis of AF and treatment with anticoagulation may prevent strokes.
Methods Using a multi-institutional EHR, this study identified 412,085 individuals 45 to 95 years of age without prevalent AF between 2000 and 2014. A prediction model was derived and validated for 5-year AF risk by using split-sample validation and model performance was compared with other methods of AF risk assessment.
Results Within 5 years, 14,334 individuals developed AF. In the derivation sample (7,216 AF events of 206,042 total), the optimal risk model included sex, age, race, smoking, height, weight, diastolic blood pressure, hypertension, hyperlipidemia, heart failure, coronary heart disease, valvular disease, prior stroke, peripheral arterial disease, chronic kidney disease, hypothyroidism, and quadratic terms for height, weight, and age. In the validation sample (7,118 AF events of 206,043 total) the AF risk model demonstrated good discrimination (C-statistic: 0.777; 95% confidence interval [CI:] 0.771 to 0.783) and calibration (0.99; 95% CI: 0.96 to 1.01). Model discrimination and calibration were superior to CHARGE-AF (Cohorts for Heart and Aging Research in Genomic Epidemiology AF) (C-statistic: 0.753; 95% CI: 0.747 to 0.759; calibration slope: 0.72; 95% CI: 0.71 to 0.74), C2HEST (Coronary artery disease / chronic obstructive pulmonary disease; Hypertension; Elderly [age ≥75 years]; Systolic heart failure; Thyroid disease [hyperthyroidism]) (C-statistic: 0.754; 95% CI: 0.747 to 0.762; calibration slope: 0.44; 95% CI: 0.43 to 0.45), and CHA2DS2-VASc (Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Prior stroke, transient ischemic attack [TIA], or thromboembolism, Vascular disease, Age 65–74 years, Sex category [female]) scores (C-statistic: 0.702; 95% CI: 0.693 to 0.710; calibration slope: 0.37; 95% CI: 0.36 to 0.38). AF risk discriminated incident stroke (n = 4,814; C-statistic: 0.684; 95% CI: 0.677 to 0.692) and stroke within 90 days of incident AF (n = 327; C-statistic: 0.789; 95% CI: 0.764 to 0.814).
Conclusions A model developed from a real-world EHR database predicted AF accurately and stratified stroke risk. Incorporating AF prediction into EHRs may enable risk-guided screening for AF.
Footnotes
↵∗ Drs. Hulme and Khurshid contributed equally to this work and are joint first authors.
Supported by Doris Duke Charitable Foundation Clinical Research Mentorship grant 2016077 to Drs. Hulme and Lubitz, grant 2017039 to Drs. Wang and Lubitz, and Doris Duke Charitable Foundation Clinical Scientist Development Award 2014105 to Dr. Lubitz. Additional support provided by U.S. National Institutes of Health grants R01HL139731 (to Dr. Lubitz), 2R01HL092577 (Drs. Ellinor and Benjamin), 1R01HL128914 (Drs. Ellinor and Benjamin), 2U54HL120163 (Dr. Benjamin), R01NS103924 and K23NS086873 (Dr. Anderson), and R01HL104156 and K24HL105780 (Dr. Ellinor); by Massachusetts General Hospital Center for Genomic Medicine Catalyst Award (Dr. Anderson); by American Heart Association grant 18SFRN34250007 (Drs. Lubitz and Anderson), 18SFRN34110082 (Dr. Benjamin), 18SFRN34150007 (Dr. Trinquart); postdoctoral fellowship award 17POST33660226 (Dr. Weng); and by Established Investigator Award 13EIA14220013 (Dr. Ellinor). Dr. Anderson is a consultant for ApoPharma. Dr. Benjamin serves on the National Heart Lung Institute's CARDIA Blood Observational Safety Monitoring Board and was Associate Editor for Circulation until June 2016. Dr. Ashburner has received research support from Boehringer Ingelheim and Bristol-Myers Squibb/Pfizer. Dr. McManus has received research support and is a consultant for Bristol-Myers Squibb, Pfizer, and Flexcon; has received research support from Apple, Philips, and Boehringer Ingelheim; and is a consultant for Boston Biomedical Associates. Dr. Ellinor is a principal investigator for Bayer AG; and has received research support from Bayer AG, Novartis, and Quest Diagnostics. Dr. Lubitz has received research support from Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, and Bayer AG; and has consulted for Abbott, Bristol-Myers Squibb/Pfizer, Quest Diagnostics, and Bayer AG. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Received June 11, 2019.
- Accepted July 22, 2019.
- 2019 American College of Cardiology Foundation
This article requires a subscription or purchase to view the full text. If you are a subscriber or member, click Login or the Subscribe link (top menu above) to access this article.