Advertisement

Oral Anticoagulation Use in High-Risk Patients Is Improved by Elimination of False-Positive and Inactive Atrial Fibrillation Cases

Open AccessPublished:December 23, 2020DOI:https://doi.org/10.1016/j.amjmed.2020.11.024

      Abstract

      Background

      Multiple registries have reported that >40% of high-risk atrial fibrillation patients are not taking oral anticoagulants. The purpose of our study was to determine the presence or absence of active atrial fibrillation and CHA2DS2-VASc (Congestive heart failure, Hypertension, Age ≥75 y, Diabetes mellitus, prior Stroke [or transient ischemic attack or thromboembolism], Vascular disease, Age 65-74 y, Sex category) risk factors to accurately identify high-risk atrial fibrillation (CHA2DS2-VASc ≥2) patients requiring oral anticoagulants and the magnitude of the anticoagulant treatment gap.

      Methods

      We retrospectively adjudicated 6514 patients with atrial fibrillation documented by at least one of: billing diagnosis, electronic medical record encounter diagnosis, electronic medical record problem list, or electrocardiogram interpretation.

      Results

      After review, 4555/6514 (69.9%) had active atrial fibrillation, while 1201 had no documented history of atrial fibrillation and 758 had a history of atrial fibrillation that was no longer active. After removing the 1201 patients without a confirmed atrial fibrillation diagnosis, oral anticoagulant use in high-risk patients increased to 71.1% (P < .0001 compared with 62.9% at baseline). Oral anticoagulant use increased to 79.7% when the 758 inactive atrial fibrillation patients were also eliminated from the analysis (P < .0001 compared with baseline). In the active high-risk atrial fibrillation group, there was no significant difference in the use of oral anticoagulants between men (80.7%) and women (78.8%) with a CHA2DS2-VASc ≥2, or in women with a CHA2DS2-VASc ≥3 (79.9%).

      Conclusions

      Current registries and health system health records with unadjudicated diagnoses over-report the number of high-risk atrial fibrillation patients not taking oral anticoagulants. Expert adjudication identifies a smaller treatment gap than previously described.

      Keywords

      Clinical Significance
      • Registries and databases using unadjudicated electronic medical record diagnoses of atrial fibrillation and stroke risk factors overestimate the oral anticoagulation treatment gap.
      • Clinicians should actively adjudicate patient problem list diagnoses to minimize errors in research and quality outcomes that use such data.

      Introduction

      Worldwide, over 33 million individuals have atrial fibrillation.
      • Chugh SS
      • Havmoeller R
      • Narayanan K
      • et al.
      Worldwide epidemiology of atrial fibrillation: a global burden of disease 2010 study.
      The projected prevalence of atrial fibrillation, given the aging population and newer detection techniques, will more than double over the next 30 years.
      • Naccarelli GV
      • Varker H
      • Lin J
      • Schulman K.
      Increasing prevalence of atrial fibrillation and flutter in the United States: is it just due to the population aging?.
      ,
      • Colilla S
      • Crow A
      • Petkun W
      • Singer DE
      • Simon T
      • Liu X.
      Estimates of current and future incidence and prevalence of atrial fibrillation in the U.S. adult population.
      The incidence of all-cause stroke in atrial fibrillation patients is 5%, and ischemic strokes associated with atrial fibrillation are often more severe, more disabling, more likely to be recurrent, and more likely to cause death than stroke from other etiologies.
      • Alkhouli M
      • Friedman PA.
      Ischemic stroke risk in patient with nonvalvular atrial fibrillation JACC Review topic of the week.
      Patients with comorbidities, as measured by higher CHADS2 (Congestive heart failure, Hypertension, Age ≥75 y, Diabetes mellitus, prior Stroke [or transient ischemic attack or thromboembolism]) and CHA2DS2-VASc (Congestive heart failure, Hypertension, Age ≥75 y, Diabetes mellitus, prior Stroke [or transient ischemic attack or thromboembolism], Vascular disease, Age 65-74 y, Sex category) scores, have higher stroke, mortality, and re-hospitalization rates.
      • Lip GYH
      • Nieuwlaat R
      • Pisters R
      • Lane DA
      • Crijns HJGM
      Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation patients using a novel risk factor approach: The Euro Heart Survey on Atrial Fibrillation.
      ,
      • Naccarelli GV
      • Panaccio MP
      • Cummins G
      • Tu N.
      CHADS2 and CHA2DS2-VASc risk factors to predict first cardiovascular hospitalization among atrial fibrillation/atrial flutter patients.
      In the Stroke Prevention of Atrial Fibrillation (SPAF) trial, warfarin, with a therapeutic international normalized ratio of 2.0 to 3.0, reduced the risk of stroke, compared with placebo, by 66%.
      Stroke Prevention in Atrial Fibrillation study. Final results.
      Novel/non-vitamin-K-dependent oral anticoagulants have demonstrated similar or better efficacy in reducing stroke and systemic embolic events in large-scale prospective trials.
      • Connolly SJ
      • Ezekowitz MD
      • Yusuf S
      • et al.
      Dabigatran versus warfarin in patients with atrial fibrillation.
      • Patel MR
      • Mahaffey KW
      • Garg J
      • et al.
      Rivaroxaban versus warfarin in nonvalvular atrial fibrillation.
      • Granger CB
      • Alexander JH
      • McMurray JJ
      • et al.
      Apixaban versus warfarin in patients with atrial fibrillation.
      • Giugliano RP
      • Ruff CT
      • Braunwald E
      • et al.
      Edoxaban versus warfarin in patients with atrial fibrillation.
      In order to treat high-risk patients and avoid oral anticoagulants in low-risk patients, current guidelines recommend anticoagulant treatment in high-risk atrial fibrillation patients with CHA2DS2-VASc scores of ≥2 in men and ≥3 in women.
      • Camm AJ
      • Lip GY
      • De Caterina R
      • et al.
      2012 Focused update of the ESC guidelines for the management of atrial fibrillation: an update of the 2010 ESC guidelines for the management of atrial fibrillation-developed with the special contribution of the European Heart Rhythm Association.
      ,
      • January CT
      • Wann LS
      • Calkins H
      • et al.
      2019 AHA/ACC/HRS focused update of the 2014 AHA/ACC/HRS Guideline for the management of patients with atrial fibrillation. A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society.
      Despite this recommendation, registries and systemic reviews suggest that >40% of high-risk atrial fibrillation patients are not taking oral anticoagulants.
      • Kowey PR
      • Reiffel JA
      • Myerburg R
      • et al.
      Warfarin and aspirin use in atrial fibrillation among practicing cardiologist (from the AFFECTS Registry).
      • Go AS
      • Hylek EM
      • Borowsky LH
      • Phillips KA
      • Selby JV
      • Singer DE.
      Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study.
      • Reynolds MR
      • Shah J
      • Essebag V
      • et al.
      Patterns and predictors of warfarin use in patients with new-onset atrial fibrillation from the FRACTAL Registry.
      • Hsu JC
      • Maddox TM
      • Kennedy KF
      • et al.
      Oral anticoagulant therapy prescription in patients with atrial fibrillation across the spectrum of stroke risk: insights from the NCDR PINNACLE Registry.
      • Pokorney SD
      • Gersch BJ
      • Ahmad A
      • et al.
      Stroke prevention in atrial fibrillation: Closing the gap.
      This treatment gap has not been significantly reduced with the commercial release of the novel/non-vitamin K-dependent oral anticoagulants.
      • Marzec LB
      • Wang J
      • Shah ND
      • et al.
      Influence of direct oral anticoagulants on rates of oral anticoagulation for atrial fibrillation.
      Multiple theories exist as to the reasons for this treatment gap.
      • Pokorney SD
      • Gersch BJ
      • Ahmad A
      • et al.
      Stroke prevention in atrial fibrillation: Closing the gap.
      ,
      • Bungard TJ
      • Ghali WA
      • Teo KT
      • McAlister FA
      • Tsuyuki RT.
      Why do patients with atrial fibrillation not receive warfarin?.
      In reviewing patients diagnosed as having atrial fibrillation as part of a quality-of-care project using our Cerner electronic medical record, we observed that active disease was not present in a large number of patients labeled as having atrial fibrillation, either because of wrong diagnosis or no documented arrhythmia recurrences in the last 5 years. Several prior reports have suggested that the misdiagnosis of atrial fibrillation in the electronic medical record is common,
      • Navar-Boggan AM
      • Rymer JA
      • Piccini JP
      • et al.
      Accuracy and validation of an automated electronic algorithm to identify patients with atrial fibrillation at risk for stroke.
      • Tu K
      • Nieuwlaat R
      • Cheng SY
      • et al.
      Identifying patients with atrial fibrillation in administrative data.
      • Khurshid S
      • Keaney J
      • Ellinor PT
      • Lubitz SA.
      A simple and portable algorithm for identifying atrial fibrillation in the electronic medical record.
      • Shah RU
      • Mukherjee R
      • Zhang Y
      • et al.
      Impact of different electronic cohort definitions to identify patients with atrial fibrillation from the electronic medical record.
      • Jensen PN
      • Johnson K
      • Floyd J
      • Heckbert SR
      • Carnahan R
      • Dublin S.
      A systematic review of validated methods for identifying atrial fibrillation using administrative data.
      and varies by the criteria used. We undertook a systematic retrospective electronic medical record review to determine the presence or absence of active atrial fibrillation to accurately identify high-risk patients who should be treated with oral anticoagulants. We hypothesized that proper identification of such patients would lower the reported treatment gap of patients who should receive oral anticoagulants and more accurately define real-world practices of the use of oral anticoagulants in high-risk atrial fibrillation patients.

      Methods

      To identify patients with presumed atrial fibrillation from our electronic medical record, we used diagnosis codes from inpatient, outpatient, and emergency department visits in the Penn State Health system from 2015 thru 2017. Patients were selected if they had 2 or more outpatient visits with a cardiologist or primary care provider. Patients with atrial fibrillation were identified by screening for an International Classification of Diseases, Ninth Revision code of 427.31 (atrial fibrillation) or 427.32 (atrial flutter) coded for 1 inpatient or 2 outpatient or emergency department visits over the last 5 years. Atrial fibrillation was documented by at least one of: billing diagnosis, electronic medical record encounter diagnosis, electronic medical record problem list, or electrocardiogram interpretation.
      Based on our criteria, 6514 patients with presumed atrial fibrillation were selected and made up our study group. Figure 1 is a diagram of the patient flow. All patients had their electronic charts manually reviewed by one of the authors to determine if the diagnosis of atrial fibrillation was correct, and also to determine if the diagnosis was still active (occurrences within the last 5 years without rhythm control treatment). In active atrial fibrillation patients, part of the expert review included a recommendation of whether or not the patient should be taking an oral anticoagulant. Patients without active atrial fibrillation were subclassified into no prior history (wrong or inaccurate diagnosis) or inactive (an accurate history of atrial fibrillation that was no longer clinically active in at least the last 5 years without a rhythm control strategy). In addition, all CHA2DS2-VASc risk factors were also reviewed for accuracy utilizing data as published by van Doom et al.
      • Van Doom S
      • Brakenhoff TB
      • Moons KGM
      • et al.
      The effects of misclassification in routine healthcare databases on the accuracy of prognostic prediction models: a case study of the CHA2DS2–VASc score in atrial fibrillation.
      Based on the timing of this review, those with CHA2DS2-VASc risk scores of ≥2 were considered high-risk atrial fibrillation patients. Given changes in the guidelines, at the time of analysis of the data, further analysis of women with scores of 3 or higher are included. A CHA2DS2-VASc risk score of 0 was considered low risk and a score of 1, intermediate risk. Patients taking warfarin had their time in therapeutic range calculated using the Rosendaal method.
      • Rosendaal FR
      • Cannegieter SC
      • van der Meer FJ
      • Briet E.
      A method to determine the optimal intensity of oral anticoagulant therapy.
      Figure 1
      Figure 1Flow diagram of patient flow. AF = atrial fibrillation; HRAF = high-risk atrial fibrillation; LRAF = low-risk atrial fibrillation; OAC = oral anticoagulation use.

      Results

      Active atrial fibrillation was noted in 4555/6514 patients (69.9%) after 1959 patients (30.1%) with no prior history (n = 1201) or inactive atrial fibrillation (n = 789) were withdrawn from further analysis (Figure 1, Table 1). At baseline in the presumed atrial fibrillation group, 62.9% of high-risk atrial fibrillation patients were taking an oral anticoagulant. In the postadjudicated active atrial fibrillation group, 3882 had high-risk atrial fibrillation with CHA2DS2-VASc ≥2. Oral anticoagulant use in these postadjudication high-risk atrial fibrillation patients (3096/3882 [79.7%]) was statistically higher than oral anticoagulant use (62.9%) in the high-risk atrial fibrillation patients preadjudication (P < .0001 by McNemar test) (Figure 2A, Table 2). In the active high-risk atrial fibrillation group, there was no significant difference in the use of oral anticoagulants between men (80.7%) and women (78.8%) with a CHA2DS2-VASc ≥2, or in women with a CHA2DS2-VASc ≥3 (79.9%). The use of novel oral anticoagulants (51.7%) was numerically but not statistically higher than warfarin (48.3%). Age of 90 years or older was associated with a doubling of oral anticoagulants not being prescribed, from 20% to 40.4% (P < .05). In high-risk atrial fibrillation patients taking an oral anticoagulant, expert review agreed in 95% of the cases. In high-risk atrial fibrillation patients not taking an oral anticoagulant (20.3%), expert review disagreed with that recommendation 60% of the time, suggesting that oral anticoagulants can be taken in up to 92% of high-risk atrial fibrillation patients without an absolute contraindication or repeated patient refusal. Table 1 lists the patient characteristics of presumed atrial fibrillation diagnosis at baseline, the active atrial fibrillation group, and the false-positive atrial fibrillation groups (No atrial fibrillation documented, Inactive atrial fibrillation, and the combined false-positive group). The Active high-risk atrial fibrillation group had oral anticoagulants prescribed in 75.0% of the cases, vs only 15% in the false-positive group.
      Figure 2
      Figure 2(A) OAC use at baseline in preadjudicated AF (gray bars) and after postadjudicated AF (excluding patients with no history of AF and inactive AF) after expert review (black bars) by CHA2DS2-VASc score. There was statistically higher OAC use after expert review of active HRAF (CHA2DS2-VASc ≥2) status (P < .0001). (B) OAC use at baseline in preadjudicated AF (gray bars) and after postadjudicated AF (excluding patients with no history of AF but including patients with inactive AF) after expert review (black and white stippled bars) by CHA2DS2-VASc score. There was statistically higher OAC use after expert review of active HRAF (CHA2DS2-VASc ≥2) status (P < .0001). AF = atrial fibrillation; CHA2DS2-VASc = Congestive heart failure, Hypertension, Age ≥75 y, Diabetes mellitus, prior Stroke (or transient ischemic attack or thromboembolism), Vascular disease, Age 65-74 y, Sex category; HRAF = high-risk atrial fibrillation; OAC = oral anticoagulation use.
      Table 1Patients’ Characteristics
      All Patients Pre-Adj n = 6514Active AF n = 4555No AF + Inactive AF n = 1959NO AF n = 1201Inactive AF n = 758
      Age, mean ± SD70.8 ± 13.572.2 ± 12.767.4 ± 14.667.367.6
      Male sex, n (%)3966 (60.9%)2760 (60.6%)1206 (61.6%)677 (56.3%)529 (69.8%)
      CHA2DS2-VASc, mean ± SD3.1 ± 1.73.3 ± 1.62.6 ± 1.72.72.6
      Hypertension, n (%)4296 (66.0%)3222 (70.1%)1074 (54.8%)608 (50.6%)466 (61.5%)
      Diabetes mellitus, n (%)1606 (24.7%)1191 (26.1%)415 (21.2%)230 (19.2%)185 (24.4%)
      Stroke or TIA, n (%)821 (12.6%)611 (13.4%)210 (10.2%)138 (11.5%)72 (9.5%)
      Congestive heart failure, n (%)1534 (23.5%)1241 (27.2%)293 (15.0%)168 (14.0%)125 (16.5%)
      Vascular disease, n (%)1980 (30.4%)1414 (31.0%)566 (28.9%)297 (24.7%)269 (35.5%)
      Oral anticoagulation use, n (%)3719 (57.1%)3415 (75.0%)304 (15.5%)219 (18.2%)85 (7.1%)
      AF = atrial fibrillation; CHA2DS2-VASc = Congestive heart failure, Hypertension, Age ≥75 y, Diabetes mellitus, prior Stroke (or transient ischemic attack or thromboembolism), Vascular disease, Age 65-74 y, Sex category; SD = standard deviation; TIA = transient ischemic attack.
      Table 2Oral Anticoagulation (OAC) Use Prior to and After Adjudication in Overall and High-Risk Atrial Fibrillation (HRAF) Groups
      n%P Value
      Baseline AF OAC use3719/651457.1
      Baseline HRAF OAC use3320/527962.9
      Active AF OAC use3415/455575.0< .0001
      P value calculated by McNemar test compared with baseline.
      Active HRAF OAC use3096/388279.7< .0001
      P value calculated by McNemar test compared with baseline.
      AF minus NO AF OAC use3500/531365.9
      HRAF minus No AF OAC use3160/444671.1< .0001
      P value calculated by McNemar test compared with baseline.
      No AF history OAC use219/120118.2
      No HRAF history OAC use160/83319.2
      Inactive AF OAC use85/75811.2
      Inactive HRAF OAC use64/58510.9
      AF = atrial fibrillation.
      low asterisk P value calculated by McNemar test compared with baseline.
      A further analysis after expert adjudication was performed, including the inactive atrial fibrillation group and only removing the 1201 patients without a confirmed atrial fibrillation diagnosis (n = 5313). In this analysis, oral anticoagulant use in high-risk atrial fibrillation patients still significantly increased from 62.9% at baseline to 71.1% (P < .0001) (Figure 2B, Table 2).
      Expert review of active atrial fibrillation agreed with the problem list 85.6%, encounter diagnosis 83.9%, and positive electrocardiogram in 84.5%, but agreed only 71.4% with the billing diagnosis.
      Warfarin use was higher in the false-positive atrial fibrillation group that had another reason for oral anticoagulant use (81.3%), compared with the inactive or misdiagnosed group (60.7%; P = .0001). Overall, the time in therapeutic range control was excellent in our warfarin-treated patients, with 68% of patients having a time in the therapeutic range ≥70% using the Rosendaal method. However, this left 32% of warfarin patients with a time in therapeutic range <70%.
      In the group of patients with no documented or inactive atrial fibrillation, oral anticoagulant use was low, being used in 304/1959 (15.5%), with 161/1959 (8.2%) patients taking oral anticoagulants for indications not related to atrial fibrillation and only 143/1959 (7.3%) taking an oral anticoagulant for inactive or misdiagnosed atrial fibrillation. Thus, 143/304 (47.0%) nonatrial fibrillation patients were taking an oral anticoagulant for inactive or misdiagnosed atrial fibrillation.
      After adjudication, 183 patients had a CHA2DS2-VASc = 0 and 60/183 (32.8%) were taking an oral anticoagulant. The majority of those taking anticoagulants were patients in the pericardioversion or ablation time period. In the CHA2DS2-VASc = 1 group, 259/490 (56.4%) were taking an oral anticoagulant.
      Similar to our findings of verifying active atrial fibrillation, CHA2DS2-VASc risk factors were 10% to Table 2 18% inaccurate compared with the chart review of each patient. The positive predictive value, by expert review, of a diagnosis of stroke or transient ischemic attack was 28% lower, congestive heart failure 27% lower, hypertension 11% lower, diabetes mellitus 16% lower, and peripheral artery or coronary artery disease 20% lower than stated in the electronic medical record. Of note, only 7.3% of these corrections led to a change of score large enough to alter an oral anticoagulant recommendation.

      Discussion

      The main novel finding of our study is that current registries, insurance databases, and health system records, with unadjudicated electronic medical record diagnoses, over-report the number of high-risk atrial fibrillation patients not taking an oral anticoagulant and thus, overestimate the treatment gap in such patients. Our 30% false-positive diagnosis of atrial fibrillation was higher than previous studies,
      • Navar-Boggan AM
      • Rymer JA
      • Piccini JP
      • et al.
      Accuracy and validation of an automated electronic algorithm to identify patients with atrial fibrillation at risk for stroke.
      • Tu K
      • Nieuwlaat R
      • Cheng SY
      • et al.
      Identifying patients with atrial fibrillation in administrative data.
      • Khurshid S
      • Keaney J
      • Ellinor PT
      • Lubitz SA.
      A simple and portable algorithm for identifying atrial fibrillation in the electronic medical record.
      • Shah RU
      • Mukherjee R
      • Zhang Y
      • et al.
      Impact of different electronic cohort definitions to identify patients with atrial fibrillation from the electronic medical record.
      however, we included inactive atrial fibrillation patients with no documented atrial fibrillation in the last 5 years. Eliminating this cohort results in a false-positive atrial fibrillation of 18%, consistent with multiple previous reports. The PINNACLE AF registry,
      • Hsu JC
      • Maddox TM
      • Kennedy KF
      • et al.
      Oral anticoagulant therapy prescription in patients with atrial fibrillation across the spectrum of stroke risk: insights from the NCDR PINNACLE Registry.
      at over a hundred study sites, demonstrated a wide range (8%-89%) of proper oral anticoagulants use in atrial fibrillation patients. Dozens of sites from this registry reported anticoagulant use of <30% in high-risk atrial fibrillation patients. The cause of such low oral anticoagulant use in these patients can be explained by the fact that the electronic medical record diagnosis of high-risk atrial fibrillation from these sites was overestimated and consistent with our findings.
      By excluding patients with absolute contraindications or those who repeatedly refused to take an oral anticoagulant, we found a ceiling of oral anticoagulant use to be 92% of high-risk atrial fibrillation patients similar to the high level of anticoagulant use reported in ORBIT-AF, GARFIELD-AF, and the best performers in the PINNACLE AF Registry.
      • Hsu JC
      • Maddox TM
      • Kennedy KF
      • et al.
      Oral anticoagulant therapy prescription in patients with atrial fibrillation across the spectrum of stroke risk: insights from the NCDR PINNACLE Registry.
      ,
      • Steinberg BA
      • Gao H
      • Shrader P
      • et al.
      International trends in clinical characteristics and oral anticoagulation treatment for patients with atrial fibrillation: results from the GARFIELD-AF and ORBIT AF-II Registries.
      Thus, <10% of high-risk atrial fibrillation patients seem to have absolute contraindications to the use of anticoagulants or refuse to take this recommended therapy even after repeated shared-decision making.
      • Steinberg BA
      • Ballew NG
      • Greiner MA
      • et al.
      Ischemic and bleeding outcomes in patients with atrial fibrillation and contraindications to oral anticoagulation.
      • Overvad TF
      • Nielsen PB
      • Lip GYH.
      Treatment thresholds for stroke prevention in atrial fibrillation: observations on the CHA2DS2VASc score.
      • Seaburg L
      • Hess EP
      • Coylewright M
      • Ting HT
      • McLeod CJ
      • Montori VM.
      Shared decision making in atrial fibrillation. Where we are and where we should be going.
      Overall in our study, only 20.3% of high-risk atrial fibrillation patients were not taking an oral anticoagulant, and this number statistically doubled to 40.4% in patients 90 years or older, consistent given the frailty of this patient group. Data have shown that nuisance bleeding and falls are not associated with the feared outcomes that minimize the appropriate use of anticoagulants in high-risk atrial fibrillation patients.
      • O'Brien EC
      • Holmes DN
      • Thomas LE
      • et al.
      Prognostic significance of nuisance bleeding in anticoagulated patients with atrial fibrillation.
      ,
      • Donze J
      • Clair C
      • Hug B
      • et al.
      Risk of falls and major bleeds in patients on oral anticoagulation therapy.
      Many of the remaining patients with absolute contraindications to anticoagulant therapy may be candidates for left atrial occluding devices.
      • Holmes DR
      • Reddy VY
      • Turi ZG
      • et al.
      Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial.
      In our data, as in other studies, we took credit for anticoagulant use in all warfarin patients, independent of their time in therapeutic range. However, one-third of our warfarin-treated patients had a time in therapeutic range of <60%, and such patients should be a strategic target for a switch to non-vitamin-K-dependent oral anticoagulants.
      • Bonde AN
      • Staerk L
      • Lee CJY
      • et al.
      Outcomes among patients with atrial fibrillation and appropriate anticoagulation control.
      • Pokorney SD
      • Holmes DN
      • Thomas L
      • et al.
      Association between warfarin control metrics and atrial fibrillation outcomes in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation.
      • Wallentin L
      • Lopes RD
      • Hanna M
      • et al.
      Efficacy and safety of apixaban compared with warfarin at different levels of predicted international normalized ratio control for stroke prevention of atrial fibrillation.
      Bonde et al
      • Bonde AN
      • Staerk L
      • Lee CJY
      • et al.
      Outcomes among patients with atrial fibrillation and appropriate anticoagulation control.
      noted in a Scandinavian database that two-thirds of patients had a time in therapeutic range of <70%. and these patients had a higher stroke and major bleeding rate compared with the group who remained in the therapeutic range ≥70% of the time.
      Our study also validated the results of other studies
      • Horskey J
      • Drucker EA
      • Ramelson HZ.
      Accuracy and completeness of clinical coding using ICD-10 for ambulatory visits.
      • Bayley KB
      • Belnap T
      • Savitz L
      • Masica AL
      • Shah N
      • Fleming NS.
      Challenges in using electronic health record data for CER. Experience of 4 learning organizations and solutions applied.
      • Thigpen JL
      • Dillon C
      • Forster KB
      • et al.
      Validity of international classification of disease codes to identify ischemic stroke and intracranial hemorrhage among individuals with associated diagnosis of atrial fibrillation.
      • Blecker S
      • Katz SD
      • Horwitz LI
      • et al.
      Comparison of approaches for heart failure case identification from electronic health record data.
      • Ahmad FS
      • Chan C
      • Rosenman MB
      • et al.
      Validity of cardiovascular data from electronic sources. The Multi-Ethnic study of atherosclerosis and HealthLNK.
      that the correct diagnosis of all the risk factors may be inaccurate and affect the risk score of such patients. In our study, the inaccuracy of risk-factor diagnosis rarely (7.3%) resulted in a change of risk score large enough to alter the recommendation of whether the patients should take an anticoagulant or not. Improving the accuracy of atrial fibrillation and all risk-factor diagnoses has importance related to quality-improvement projects to improve anticoagulant use
      • Donze J
      • Clair C
      • Hug B
      • et al.
      Risk of falls and major bleeds in patients on oral anticoagulation therapy.
      ,
      • Piccini JP
      • Xu H
      • Cox M
      • et al.
      Adherence to guideline-directed stroke prevention therapy for atrial fibrillation is achievable. First results from Get With The Guidelines-Atrial-Fibrillation (GWTG-AFIB).
      and for future projects using databases to predict atrial fibrillation for further screening and interventions.
      • Vinereanu D
      • Lopes RD
      • Bahit MC
      • et al.
      A multifaceted intervention to improve treatment with oral anticoagulants in atrial fibrillation (IMPACT-AF): an international, cluster-randomised trial.
      • Piazza G
      • Hurwitz S
      • Galvin CE
      • et al.
      Alert-based computerized decision support for high-risk hospitalized patients with atrial fibrillation not prescribed anticoagulation: a randomized, controlled trial (AF-ALERT).
      • Chen YC
      • Roebuck AE
      • Sami A
      • Ersin OH
      • Mirro MJ.
      The use of electronic personal health records to improve medication adherence and patient engagement: A randomized study of non-valvular atrial fibrillation patients.
      • Holt TA
      • Fitzmaurice DA
      • Marshall T
      • et al.
      Automated risk assessment for stroke in atrial fibrillation (AURAS-AF) – an automated software system to promote anticoagulation and reduce stroke risk: study protocol for a cluster randomized trial.
      • Kolek MJ
      • Graves AJ
      • Xu M
      • et al.
      Evaluation of a prediction model for the development of atrial fibrillation in a repository of electronic medical records.
      • Lively SH
      • DeRemer CE
      • Carroll M
      • Maddox WR
      • Johnson MH
      • Berman AE.
      Electronic medical record (EMR)-derived appropriate inpatient risk stratification and anticoagulation treatment of atrial fibrillation at a large academic medical center: opportunities for EMR-based clinical decision-making support tools.
      • Wang SV
      • Rogers JR
      • Jin Y
      • Bates DW
      • Fischer MA.
      Use of electronic healthcare records to identify complex patients with atrial fibrillation for targeted intervention.
      • Hulme OL
      • Khurshid S
      • Weng LC
      • et al.
      Development and validation of a prediction model for atrial fibrillation using electronic health records.
      We noted no sex difference in appropriate anticoagulant use in patients, similar to some other studies.
      • Loikas D
      • Forslund T
      • Wettermark B
      • Schenck-Gustafson K
      • Hjemdahl P
      • von Euler M
      Sex and gender differences in thromboprophylactic treatment of patients with atrial fibrillation after the introduction of non-vitamin K oral anticoagulants.
      ,
      • Piccini JP
      • Simon DN
      • Steinberg BA
      • et al.
      Differences in clinical and functional outcomes of atrial fibrillation in women and men: two-year results from the ORBIT-AF registry.
      Loikas et al
      • Loikas D
      • Forslund T
      • Wettermark B
      • Schenck-Gustafson K
      • Hjemdahl P
      • von Euler M
      Sex and gender differences in thromboprophylactic treatment of patients with atrial fibrillation after the introduction of non-vitamin K oral anticoagulants.
      noted that anticoagulant use in women improved and was similar to men by 2015, compared with a lower anticoagulant use in women in 2011. This time-period was when the CHA2DS2-VASc scoring system replaced CHADS2, and non-vitamin-K-dependent oral anticoagulants were introduced. Piccini et al
      • Piccini JP
      • Simon DN
      • Steinberg BA
      • et al.
      Differences in clinical and functional outcomes of atrial fibrillation in women and men: two-year results from the ORBIT-AF registry.
      also noted no sex difference in oral anticoagulant use in high-risk atrial fibrillation patients in the 2010-2012 timeframe.
      Our study found that expert review was concordant with the decision of the practitioner to prescribe an anticoagulant in over 90% of cases. The majority of expert review disagreement was noted in high-risk atrial fibrillation not prescribed an anticoagulant based on perceived risk of bleeding, age, frailty, and falls. Recent registry data
      • O'Brien EC
      • Holmes DN
      • Thomas LE
      • et al.
      Prognostic significance of nuisance bleeding in anticoagulated patients with atrial fibrillation.
      noted that over 96% of patients who had minor nuisance bleeding issues continued on oral anticoagulants without any increase in major bleeds.
      Based on our findings, all diagnoses may have similar inaccuracies and highlight the limitations of studies with any diagnosis that use unadjudicated Marketscan, Medicare, Department of Defense, insurance, and other electronic medical record databases. Along with an 18% error rate in identifying a diagnosis of atrial fibrillation, other risk factors had a 10% to 18% inaccuracy rate. Prospective registries, such as ORBIT-AF and Garfield AF,
      • Steinberg BA
      • Gao H
      • Shrader P
      • et al.
      International trends in clinical characteristics and oral anticoagulation treatment for patients with atrial fibrillation: results from the GARFIELD-AF and ORBIT AF-II Registries.
      can minimize gaps in data and inaccuracies by correction of the data, based on inquiries of the monitoring board.
      There are some limitations of our study. Does expert review represent a true gold standard? There is the possibility of inter-reviewer variability. Expert review would not be accurate related to silent occurrences or from atrial fibrillation episodes not documented in our electronic medical record. We tried to minimize this issue by only including patients that frequented our health system so that our database was reasonably complete, including outside scanned records. However, patients may have had silent recurrences, and this may be more of a limitation in the inactive atrial fibrillation group.
      Selecting sensitive criteria for entry was done on purpose to minimize missing patients and thus, the loss of some specificity was not surprising. The criteria used in this study are similar to most other studies for enrolling atrial fibrillation patients in registries and clinical trials. As noted above, the magnitude of false positive was higher than previous studies, but this is partially explained by our study including an inactive atrial fibrillation group not included in other prior reports. Even with this wide entry criteria to identify patients, some atrial fibrillation patients were still not identified.
      One could argue that our data might be site specific or electronic medical record vendor specific. However, a more likely explanation is that providers who accurately update problem list diagnoses provide the most accurate data, independent of the manual or electronic medical record used. Because the problem list can be edited after the fact, future attempts to accurately define specific diagnoses should concentrate on correcting active problem lists because encounter and billing diagnoses cannot be edited after the fact.
      Our study did not exclude the following patients: mechanical valves, postcardiac transplant, adult congenital heart disease, or patients who had undergone an ablation procedure. High-risk atrial fibrillation patients, post ablation procedure, were still considered to need oral anticoagulants as per current recommendations. Our study did not analyze the small number of patients who have indications for oral anticoagulants outside the CHA2DS2-VASc scoring system, such as hypertrophic cardiomyopathy or hyperthyroidism. The effect of our findings in subgroups of permanent, persistent, or paroxysmal atrial fibrillation and rate vs rhythm treatment was not analyzed.
      Our findings help explain the over-reported underuse of oral anticoagulants in high-risk atrial fibrillation patients and partially explain that even with the approval of the novel anticoagulants, the presumed treatment gap stayed large. Prior reports of oral anticoagulant underuse based on actual or perceived bleeding risk, misperceptions about the efficacy/safety of warfarin, overestimation of aspirin's effectiveness in preventing stroke, adherence issues, especially after stopping an anticoagulant for a procedure and guideline changes over the years, may further explain the current treatment gap. Although our adjudication demonstrated a smaller anticoagulant treatment gap in high-risk atrial fibrillation patients, providing accurate electronic medical record alerts to providers of such patients and developing interventions to improve the education of patients may further decrease the treatment gap in such patients.
      • Vinereanu D
      • Lopes RD
      • Bahit MC
      • et al.
      A multifaceted intervention to improve treatment with oral anticoagulants in atrial fibrillation (IMPACT-AF): an international, cluster-randomised trial.
      • Piazza G
      • Hurwitz S
      • Galvin CE
      • et al.
      Alert-based computerized decision support for high-risk hospitalized patients with atrial fibrillation not prescribed anticoagulation: a randomized, controlled trial (AF-ALERT).
      • Chen YC
      • Roebuck AE
      • Sami A
      • Ersin OH
      • Mirro MJ.
      The use of electronic personal health records to improve medication adherence and patient engagement: A randomized study of non-valvular atrial fibrillation patients.
      • Holt TA
      • Fitzmaurice DA
      • Marshall T
      • et al.
      Automated risk assessment for stroke in atrial fibrillation (AURAS-AF) – an automated software system to promote anticoagulation and reduce stroke risk: study protocol for a cluster randomized trial.
      High-risk atrial fibrillation patients, who have major bleeding contraindications or treatment failures with anticoagulants should be screened for other therapies such as left atrial occluding devices.
      • Holmes DR
      • Reddy VY
      • Turi ZG
      • et al.
      Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial.
      Finally, accurate diagnosis of active atrial fibrillation patients will help improve future predictive models.
      • Kolek MJ
      • Graves AJ
      • Xu M
      • et al.
      Evaluation of a prediction model for the development of atrial fibrillation in a repository of electronic medical records.
      • Lively SH
      • DeRemer CE
      • Carroll M
      • Maddox WR
      • Johnson MH
      • Berman AE.
      Electronic medical record (EMR)-derived appropriate inpatient risk stratification and anticoagulation treatment of atrial fibrillation at a large academic medical center: opportunities for EMR-based clinical decision-making support tools.
      • Wang SV
      • Rogers JR
      • Jin Y
      • Bates DW
      • Fischer MA.
      Use of electronic healthcare records to identify complex patients with atrial fibrillation for targeted intervention.
      • Hulme OL
      • Khurshid S
      • Weng LC
      • et al.
      Development and validation of a prediction model for atrial fibrillation using electronic health records.

      Conclusions

      Current registries with unadjudicated electronic medical record diagnoses over-report the number of high-risk atrial fibrillation patients not taking oral anticoagulants. Expert adjudication of an active atrial fibrillation diagnosis identifies a significantly smaller oral anticoagulant treatment gap than previously reported. This issue can be minimized by individual review of patients’ medical records and correction of the problem list for the diagnosis of active atrial fibrillation and CHA2DS2-VASc risk factors.

      References

        • Chugh SS
        • Havmoeller R
        • Narayanan K
        • et al.
        Worldwide epidemiology of atrial fibrillation: a global burden of disease 2010 study.
        Circulation. 2014; 129: 837-847
        • Naccarelli GV
        • Varker H
        • Lin J
        • Schulman K.
        Increasing prevalence of atrial fibrillation and flutter in the United States: is it just due to the population aging?.
        Am J Cardiol. 2009; 104: 1534-1539
        • Colilla S
        • Crow A
        • Petkun W
        • Singer DE
        • Simon T
        • Liu X.
        Estimates of current and future incidence and prevalence of atrial fibrillation in the U.S. adult population.
        Am J Cardiol. 2013; 112: 1142-1147
        • Alkhouli M
        • Friedman PA.
        Ischemic stroke risk in patient with nonvalvular atrial fibrillation JACC Review topic of the week.
        J Am Coll Cardiol. 2019; 74: 3050-3065
        • Lip GYH
        • Nieuwlaat R
        • Pisters R
        • Lane DA
        • Crijns HJGM
        Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation patients using a novel risk factor approach: The Euro Heart Survey on Atrial Fibrillation.
        Chest. 2010; 137: 263-272
        • Naccarelli GV
        • Panaccio MP
        • Cummins G
        • Tu N.
        CHADS2 and CHA2DS2-VASc risk factors to predict first cardiovascular hospitalization among atrial fibrillation/atrial flutter patients.
        Am J Cardiol. 2012; 109: 1526-1533
      1. Stroke Prevention in Atrial Fibrillation study. Final results.
        Circulation. 1991; 84: 527-539
        • Connolly SJ
        • Ezekowitz MD
        • Yusuf S
        • et al.
        Dabigatran versus warfarin in patients with atrial fibrillation.
        N Engl J Med. 2009; 361: 1139-1151
        • Patel MR
        • Mahaffey KW
        • Garg J
        • et al.
        Rivaroxaban versus warfarin in nonvalvular atrial fibrillation.
        N Engl J Med. 2011; 365: 883-891
        • Granger CB
        • Alexander JH
        • McMurray JJ
        • et al.
        Apixaban versus warfarin in patients with atrial fibrillation.
        N Engl J Med. 2011; 15: 981-992
        • Giugliano RP
        • Ruff CT
        • Braunwald E
        • et al.
        Edoxaban versus warfarin in patients with atrial fibrillation.
        N Engl J Med. 2013; 369: 2093-2104
        • Camm AJ
        • Lip GY
        • De Caterina R
        • et al.
        2012 Focused update of the ESC guidelines for the management of atrial fibrillation: an update of the 2010 ESC guidelines for the management of atrial fibrillation-developed with the special contribution of the European Heart Rhythm Association.
        Eur Heart J. 2012; 33: 2719-2747
        • January CT
        • Wann LS
        • Calkins H
        • et al.
        2019 AHA/ACC/HRS focused update of the 2014 AHA/ACC/HRS Guideline for the management of patients with atrial fibrillation. A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society.
        J Am Coll Cardiol. 2019; 74: 104-132
        • Kowey PR
        • Reiffel JA
        • Myerburg R
        • et al.
        Warfarin and aspirin use in atrial fibrillation among practicing cardiologist (from the AFFECTS Registry).
        Am J Cardiol. 2010; 105: 1130-1134
        • Go AS
        • Hylek EM
        • Borowsky LH
        • Phillips KA
        • Selby JV
        • Singer DE.
        Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study.
        Ann Intern Med. 1999; 131: 927-934
        • Reynolds MR
        • Shah J
        • Essebag V
        • et al.
        Patterns and predictors of warfarin use in patients with new-onset atrial fibrillation from the FRACTAL Registry.
        Am J Cardiol. 2006; 97: 538-543
        • Hsu JC
        • Maddox TM
        • Kennedy KF
        • et al.
        Oral anticoagulant therapy prescription in patients with atrial fibrillation across the spectrum of stroke risk: insights from the NCDR PINNACLE Registry.
        JAMA Cardiol. 2016;; 1: 55-62
        • Pokorney SD
        • Gersch BJ
        • Ahmad A
        • et al.
        Stroke prevention in atrial fibrillation: Closing the gap.
        Am Heart J. 2019; 210: 29-38
        • Marzec LB
        • Wang J
        • Shah ND
        • et al.
        Influence of direct oral anticoagulants on rates of oral anticoagulation for atrial fibrillation.
        J Am Coll Cardiol. 2017; 69: 2475-2484
        • Bungard TJ
        • Ghali WA
        • Teo KT
        • McAlister FA
        • Tsuyuki RT.
        Why do patients with atrial fibrillation not receive warfarin?.
        Arch Intern Med. 2000; 160: 41-46
        • Navar-Boggan AM
        • Rymer JA
        • Piccini JP
        • et al.
        Accuracy and validation of an automated electronic algorithm to identify patients with atrial fibrillation at risk for stroke.
        Am Heart J. 2015; 169: 39-44.e2
        • Tu K
        • Nieuwlaat R
        • Cheng SY
        • et al.
        Identifying patients with atrial fibrillation in administrative data.
        Can J Cardiol. 2016; 32: 1561-1565
        • Khurshid S
        • Keaney J
        • Ellinor PT
        • Lubitz SA.
        A simple and portable algorithm for identifying atrial fibrillation in the electronic medical record.
        Am J Cardiol. 2016; 117: 221-225
        • Shah RU
        • Mukherjee R
        • Zhang Y
        • et al.
        Impact of different electronic cohort definitions to identify patients with atrial fibrillation from the electronic medical record.
        J Am Heart Assoc. 2020; 9e014527
        • Jensen PN
        • Johnson K
        • Floyd J
        • Heckbert SR
        • Carnahan R
        • Dublin S.
        A systematic review of validated methods for identifying atrial fibrillation using administrative data.
        Pharmacoepidiol Drug Saf. 2012; 21: 141-147
        • Van Doom S
        • Brakenhoff TB
        • Moons KGM
        • et al.
        The effects of misclassification in routine healthcare databases on the accuracy of prognostic prediction models: a case study of the CHA2DS2–VASc score in atrial fibrillation.
        Diagn Progn Res. 2017; 1: 18
        • Rosendaal FR
        • Cannegieter SC
        • van der Meer FJ
        • Briet E.
        A method to determine the optimal intensity of oral anticoagulant therapy.
        Thromb Haemost. 1993; 69: 236-239
        • Steinberg BA
        • Gao H
        • Shrader P
        • et al.
        International trends in clinical characteristics and oral anticoagulation treatment for patients with atrial fibrillation: results from the GARFIELD-AF and ORBIT AF-II Registries.
        Am Heart J. 2017; 194: 132-140
        • Steinberg BA
        • Ballew NG
        • Greiner MA
        • et al.
        Ischemic and bleeding outcomes in patients with atrial fibrillation and contraindications to oral anticoagulation.
        JACC Clin Electrophysiol. 2019; 5: 1384-1392
        • Overvad TF
        • Nielsen PB
        • Lip GYH.
        Treatment thresholds for stroke prevention in atrial fibrillation: observations on the CHA2DS2VASc score.
        Eur Heart J Cardiovasc Pharmacother. 2017; 3: 37-41
        • Seaburg L
        • Hess EP
        • Coylewright M
        • Ting HT
        • McLeod CJ
        • Montori VM.
        Shared decision making in atrial fibrillation. Where we are and where we should be going.
        Circulation. 2014; 129: 704-710
        • O'Brien EC
        • Holmes DN
        • Thomas LE
        • et al.
        Prognostic significance of nuisance bleeding in anticoagulated patients with atrial fibrillation.
        Circulation. 2018; 138: 889-897
        • Donze J
        • Clair C
        • Hug B
        • et al.
        Risk of falls and major bleeds in patients on oral anticoagulation therapy.
        Am J Med. 2012; 125: 773-778
        • Holmes DR
        • Reddy VY
        • Turi ZG
        • et al.
        Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial.
        Lancet. 2009; 374: 534-542
        • Bonde AN
        • Staerk L
        • Lee CJY
        • et al.
        Outcomes among patients with atrial fibrillation and appropriate anticoagulation control.
        J Am Coll Cardiol. 2018; 72: 1357-1365
        • Pokorney SD
        • Holmes DN
        • Thomas L
        • et al.
        Association between warfarin control metrics and atrial fibrillation outcomes in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation.
        JAMA Cardiol. 2019; 4: 756-764
        • Wallentin L
        • Lopes RD
        • Hanna M
        • et al.
        Efficacy and safety of apixaban compared with warfarin at different levels of predicted international normalized ratio control for stroke prevention of atrial fibrillation.
        Circulation. 2013; 127: 2166-2176
        • Horskey J
        • Drucker EA
        • Ramelson HZ.
        Accuracy and completeness of clinical coding using ICD-10 for ambulatory visits.
        AMIA Annu Symp Proc. 2017; 2017: 912-920
        • Bayley KB
        • Belnap T
        • Savitz L
        • Masica AL
        • Shah N
        • Fleming NS.
        Challenges in using electronic health record data for CER. Experience of 4 learning organizations and solutions applied.
        Med Care. 2013; 51: S80-S86
        • Thigpen JL
        • Dillon C
        • Forster KB
        • et al.
        Validity of international classification of disease codes to identify ischemic stroke and intracranial hemorrhage among individuals with associated diagnosis of atrial fibrillation.
        Circ Cardiovasc Qual Outcomes. 2015; 8: 8-14
        • Blecker S
        • Katz SD
        • Horwitz LI
        • et al.
        Comparison of approaches for heart failure case identification from electronic health record data.
        JAMA Cardiol. 2016; 1: 1014-1020
        • Ahmad FS
        • Chan C
        • Rosenman MB
        • et al.
        Validity of cardiovascular data from electronic sources. The Multi-Ethnic study of atherosclerosis and HealthLNK.
        Circulation. 2017; 136: 1207-1216
        • Piccini JP
        • Xu H
        • Cox M
        • et al.
        Adherence to guideline-directed stroke prevention therapy for atrial fibrillation is achievable. First results from Get With The Guidelines-Atrial-Fibrillation (GWTG-AFIB).
        Circulation. 2019; 139: 1497-1506
        • Vinereanu D
        • Lopes RD
        • Bahit MC
        • et al.
        A multifaceted intervention to improve treatment with oral anticoagulants in atrial fibrillation (IMPACT-AF): an international, cluster-randomised trial.
        Lancet. 2017; 390: 1737-1746
        • Piazza G
        • Hurwitz S
        • Galvin CE
        • et al.
        Alert-based computerized decision support for high-risk hospitalized patients with atrial fibrillation not prescribed anticoagulation: a randomized, controlled trial (AF-ALERT).
        Eur Heart J. 2020; 41: 1086-1096
        • Chen YC
        • Roebuck AE
        • Sami A
        • Ersin OH
        • Mirro MJ.
        The use of electronic personal health records to improve medication adherence and patient engagement: A randomized study of non-valvular atrial fibrillation patients.
        J Innov Card Rhythm Manag. 2017; 8: 2804-2813
        • Holt TA
        • Fitzmaurice DA
        • Marshall T
        • et al.
        Automated risk assessment for stroke in atrial fibrillation (AURAS-AF) – an automated software system to promote anticoagulation and reduce stroke risk: study protocol for a cluster randomized trial.
        Trials. 2013; 14: 385
        • Kolek MJ
        • Graves AJ
        • Xu M
        • et al.
        Evaluation of a prediction model for the development of atrial fibrillation in a repository of electronic medical records.
        JAMA Cardiol. 2016; 1: 1007-1013
        • Lively SH
        • DeRemer CE
        • Carroll M
        • Maddox WR
        • Johnson MH
        • Berman AE.
        Electronic medical record (EMR)-derived appropriate inpatient risk stratification and anticoagulation treatment of atrial fibrillation at a large academic medical center: opportunities for EMR-based clinical decision-making support tools.
        J Innov Card Rhythm Manag. 2016; 7: 2446-2450
        • Wang SV
        • Rogers JR
        • Jin Y
        • Bates DW
        • Fischer MA.
        Use of electronic healthcare records to identify complex patients with atrial fibrillation for targeted intervention.
        J Am Med Inform Assoc. 2017; 24: 339-344
        • Hulme OL
        • Khurshid S
        • Weng LC
        • et al.
        Development and validation of a prediction model for atrial fibrillation using electronic health records.
        JACC Clin Electrophysiol. 2019; 5: 1331-1341
        • Loikas D
        • Forslund T
        • Wettermark B
        • Schenck-Gustafson K
        • Hjemdahl P
        • von Euler M
        Sex and gender differences in thromboprophylactic treatment of patients with atrial fibrillation after the introduction of non-vitamin K oral anticoagulants.
        Am J Cardiol. 2017; 120: 1302-1308
        • Piccini JP
        • Simon DN
        • Steinberg BA
        • et al.
        Differences in clinical and functional outcomes of atrial fibrillation in women and men: two-year results from the ORBIT-AF registry.
        JAMA Cardiol. 2016; 1: 282-291