Abstract
Aims
There is conflicting evidence on whether the type of atrial fibrillation (AF) is associated with risk of cardiovascular events, including acute myocardial infarction (MI) and ischemic stroke. The aim of the present study was to investigate whether the risk of MI and ischemic stroke differs between individuals with first-diagnosed paroxysmal vs. non-paroxysmal AF treated with anticoagulants.
Methods and results
De-identified electronic medical records from the TriNetX federated research network were used. Individuals with a new diagnosis of paroxysmal AF who had no evidence of other types of AF in their records were 1:1 propensity score-matched with individuals with non-paroxysmal AF, defined as persistent or chronic AF, who had no evidence of other types of AF in their records. All patients were followed for three years for the outcomes of MI and ischemic stroke. Cox proportional hazard models were used to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). In the propensity-matched cohort, among 24 848 well-matched AF individuals [mean age 74.4 ± 10.4; 10 101 (40.6%) female], 410 (1.7%) were diagnosed with acute MI and 875 (3.5%) with ischemic stroke during the three-year follow-up. Individuals with paroxysmal AF had significantly higher risk of acute MI (HR: 1.65, 95%CI: 1.35–2.01) compared to those with non-paroxysmal AF. First diagnosed paroxysmal AF was associated with higher risk of non-ST elevation MI (nSTEMI) (HR: 1.89, 95%CI: 1.44–2.46). No significant association was observed between the type of AF and risk of ischemic stroke (HR: 1.09, 95%CI: 0.95–1.25).
Conclusion
Patients with first-diagnosed paroxysmal AF had higher risk of acute MI compared to individuals with non-paroxysmal AF, attributed to the higher risk of nSTEMI among patients with first-diagnosed paroxysmal AF. There was no significant association between type of AF and risk of ischemic stroke.
Keywords: Atrial fibrillation, Paroxysmal, Permanent, Myocardial infarction, Ischemic stroke
Graphical Abstract
Graphical Abstract.
What’s New.
There is no clear association of myocardial infarction (MI) or ischemic stroke with the type of atrial fibrillation (AF).
Patients with first-diagnosed paroxysmal AF had higher risk of acute MI compared to those with non-paroxysmal AF.
First-diagnosed AF was associated with higher risk of nSTEMI compared to non-paroxysmal AF, whereas the risk of STEMI was similar among patients with paroxysmal AF and non-paroxysmal AF.
There was no association between the type of AF and risk of ischemic stroke.
Introduction
Recently, the management of individuals with atrial fibrillation (AF) has evolved to better characterise the arrhythmia1 and provide integrated care management, including anticoagulation, rhythm control, and management of concomitant cardiovascular comorbidities and risk factors.2,3 Despite optimal holistic AF treatment,4 some individuals will experience new cardiovascular events, including myocardial infarction (MI) and ischemic stroke.5,6 Nonetheless, adherence to an integrated care management approach is associated with improved clinical outcomes.7
Several studies have proposed possible differences in the incidence of cardiovascular outcomes in individuals with paroxysmal AF compared to those with persistent or permanent AF. In patients with coronary artery disease undergoing coronary stenting, the presence of paroxysmal AF was associated with higher risk of recurrent MI, compared to non-paroxysmal.8 In an exploratory analysis of the EdoxabaN vs. wafarin in subjectS UndeRgoing cardiovErsion of Atrial Fibrillation (ENSURE-AF) trial, it was observed that among 2199 patients after electrical cardioversion, MI was more common in patients with paroxysmal AF compared to those with persistent during a follow-up period of 58 days (1% vs. 0.1%).9 Conversely, a prospective observational study including 3843 AF patients showed that after a median (IQR) follow-up of 3.0 (1.9–4.2) years, both stroke or systemic embolism and MI rates were similar in patients with paroxysmal AF compared to those with persistent and permanent [adjusted hazard ratio (HR):0.91, 95%CI: 0.48–1.72 and adjHR:0.95, 95%CI:0.56–1.59].10 The early treatment of AF for stroke prevention trial (EAST-AFNET 4) showed that among 1048 patients with first AF episode, after a mean follow-up period of 5.1 years, experienced higher rates of hospitalisation for acute coronary events, compared to those with previous diagnosed paroxysmal or persistent AF.11
These contradictory findings underline the need for further evidence to investigate the hypothesis that recently diagnosed paroxysmal AF may be associated with higher risk of MI compared to recently diagnosed non-paroxysmal AF. Therefore, the aim of the current study was to assess whether first-diagnosed paroxysmal AF among patients treated with oral anticoagulants (OACs) is associated with higher risk of acute MI and ischemic stroke compared to non-paroxysmal AF, using real-world data from a global federated health network.
Methods
Data availability statement and ethical approval
TriNetx is a research network utilized for several scientific purposes, compliant with the Health Insurance Portability and Accountability Act and the US federal law which protects the privacy and security of healthcare data, including de-identified data as per the de-identification standard of the HIPAA Privacy Rule (https://trinetx.com/real-world-resources/publications/). To comply with legal frameworks and ethical guidelines guarding against data re-identification, the identity of participating Healthcare Organizations (HCOs) and their individual contribution to each dataset are not disclosed. A typical HCO is a large academic health center with data coming from majority of its affiliates. A single HCO frequently has more than one facility, including main and satellite hospitals. The data are stored on the TriNetX database via a physical server at the institution’s data centre or a virtual hosted appliance. The TriNetX platform comprises of a series of these appliances connected into a federated network. This network can broadcast queries to each appliance. The network members are mainly in the United States, with an additional six HCOs in other countries that include the United Kingdom, Germany, Italy, Israel, and Singapore.12
The TriNetX platform only uses aggregated counts and statistical summaries of de-identified information. No protected health information or personal data is made available to the users of the platform. Studies using the TriNetX health research network do not need ethical approval as no patient identifiable information is received.
To gain access to the data in the TriNetX research network, requests are directed to TriNetX and a data sharing agreement is required.
Study design
This was a retrospective observational study conducted within TriNetX, which is a global federated health research network with access to electronic medical records (EMRs) from participating HCOs including academic medical centres, specialty physician practices, and community hospitals covering approximately 69.8 million individuals, from the United States, United Kingdom, Germany, Italy, Israel, and Singapore. Within this network, available data include demographics, diagnoses using International Classification of Diseases, Ninth Revision and Tenth Revision, Clinical Modification (ICD-10-CM) codes, and medications. Further information can be found online (https://trinetx.com/company-overview/). Details related to TriNetX database can be found in Supplemental material.
Cohort
The searches on the TriNetX online research platform were performed on 30 March 2023 for individuals with new diagnosis of paroxysmal, persistent or chronic AF (ICD-10-CM code I48.0, I48.1, or I48.2, respectively) who were treated with any OAC within one month after the diagnosis of AF and were followed-up for 3 years or until death. At the time of the search, there were 66 participating HCOs that returned data on the platform based on the search parameters described above (59% US, 2% ex-US, and 39% from other countries). The baseline index event date was the date that a patient was diagnosed for the first time with either paroxysmal or non-paroxysmal AF.
The cohort was divided into two groups using electronic health records according to the type of AF: individuals with new diagnosis of paroxysmal AF (ICD-10-CM code I48.0) who had no evidence of other types of AF in their records and individuals with non-paroxysmal AF, [corresponding to persistent (defined as AF that does not terminate within seven days, or that requires repeat pharmacological or electrical cardioversion:ICD-10-CM code I48.1) or chronic AF (defined as either permanent or unspecified, chronic AF which refers to any persistent, longstanding persistent or permanent AF: ICD-10-CM code I48.2)], who had no evidence of other types of AF in their records. Individuals with diagnoses of typical AF, atypical AF, and unspecified AF or atrial flutter were excluded from the analysis (ICD-10-CM code I48.3, I48.4, and I48.9, respectively).
Outcome
The outcomes of interest were the diagnosis of acute MI [captured with the ICD-10-CM code I21 in the database, including ST-elevation MI (STEMI), non ST-elevation MI (nSTEMI), and unspecified MI] and ischemic stroke (ICD-10-CM code I63) during a follow-up period of 3 years. Two exploratory analyses were performed: (i) to investigate the association of AF type with the outcomes of STEMI and nSTEMI and (ii) excluding individuals with history of acute MI and ischemic stroke before the diagnosis of AF were performed to investigate the association of AF type with the occurrence of incident acute MI and ischemic stroke.
Statistical analysis
All statistical analyses were performed on the TriNetX online research platform. Baseline characteristics were compared using chi-squared tests for categorical variables and independent-sample t-tests for continuous variables. We performed 1:1 propensity score matching to create balanced cohorts. The following variables were included in the propensity score matching: age, sex, ethnicity, co-morbidities (heart failure, arterial hypertension, ischemic stroke, dyslipidemia, diabetes mellitus, chronic kidney disease, chronic obstructive pulmonary disease, sleep apnoea, and overweight), concomitant medication (lipid-lowering treatments, beta blockers, ACE inhibitors, angiotensin II inhibitors, prior use of antiplatelet, insulin, metformin), and cholesterol levels (low-density lipoprotein cholesterol and total cholesterol) (see Supplementary material online, Table S1). After propensity score matching, Cox-proportional hazard models were used to calculate HRs and 95% confidence intervals (CIs) to assess the associations between the type of AF (i.e. first-diagnosed paroxysmal vs. non-paroxysmal) and the outcome of interest. Participants were censored when they experienced the outcome of interest or when they died. All analyses were performed in the TriNetX platform which uses R’s survival package v3.2–3.
Results
Cohort characteristics before and after propensity score matching
The initial cohort consisted of 39 290 individuals with AF who were treated with OACs (91% from the United States, 2% from Ex-US regions and 7% from unknown regions). Of these, 26 558 (67.5%) had paroxysmal AF and 12 732 (32.5%) had non-paroxysmal AF (from 53 and 54 health care organisations, respectively). Baseline characteristics of the included patients before propensity score matching are presented in Supplementary material online, Table. After propensity score matching on a 1:1 ratio, the final cohort consisted of 24 848 well-matched AF individuals (mean age 74.4 ± 10.4; 10 101 (40.6%) female) divided in two groups of 12 424, followed for 3 years. The baseline characteristics of the propensity-score matched cohort are summarized in Table 1.
Table 1.
Baseline characteristics of individuals with paroxysmal and non-paroxysmal atrial fibrillation (AF) in the propensity-score matched analysis
| Paroxysmal AF | Non-paroxysmal AF | P-value | |
|---|---|---|---|
| n = 12 428 | n = 12 428 | ||
| Demographics | |||
| Age, years (±SD) | 74.3 (±10.2) | 74.5 (±10.5) | 0.070 |
| Female, n (%) | 5037 (40.5) | 5064 (40.8) | 0.727 |
| Comorbidities | |||
| Heart failure, n (%) | 1257 (10.1) | 1348 (10.8) | 0.060 |
| Arterial hypertension, n (%) | 2876 (23.1) | 2986 (24) | 0.100 |
| Ischemic heart disease, n (%) | 1539 (12.4) | 1639 (13.2) | 0.057 |
| Ischemic Stroke, n (%) | 358 (2.9) | 393 (3.2) | 0.195 |
| Dyslipidaemia, n (%) | 1960 (15.8) | 2018 (16.2) | 0.316 |
| Diabetes mellitus, n (%) | 1135 (9.1) | 1203 (9.7) | 0.140 |
| Chronic kidney disease, n (%) | 642 (5.2) | 680 (5.5) | 0.195 |
| Overweight/obese, n (%) | 532 (4.3) | 540 (4.3) | 0.182 |
| Chronic obstructive pulmonary disease, n (%) | 495 (4.0) | 537 (4.3) | 0.182 |
| Sleep apnoea, n (%) | 389 (3.1) | 399 (3.2) | 0.717 |
| Concomitant medications and lipid levels | |||
| Lipid-lowering treatment, n (%) | 2067 (16.6) | 2123 (1.1) | 0.343 |
| Beta blockers, n (%) | 2393 (19.3) | 2485 (20) | 0.142 |
| ACE inhibitors, n (%) | 1145 (9.2) | 1176 (9.5) | 0.499 |
| Angiotensin II inhibitors, n (%) | 820 (6.6) | 826 (6.6) | 0.878 |
| Insulin, n (%) | 823 (6.6) | 849 (6.8) | 0.510 |
| Metformin, n (%) | 483 (3.9) | 454 (3.7) | 0.334 |
| Antiplatelet, n (%) | 2031 (16.3) | 2088 (16.8) | 0.331 |
| LDL cholesterol, mg/dL (±SD) | 86.4 (±33.8) | 81.5 (±33.3) | <0.001 |
AF, atrial fibrillation; SD, standard deviation; ACE, angiotensin converting enzyme; LDL, low-density lipoprotein
Type of AF and risk of mi
After propensity-score matching, among 24 848 individuals, 410 (1.7%) were diagnosed with an acute MI within 3 years of AF diagnosis, of which 260 (2.1%) occurred in individuals with paroxysmal AF and 150 (1.2%) in individuals with non-paroxysmal AF (HR: 1.65, 95%CI: 1.35–2.01, Supplementary material online, Figure SA) (Table 2). In the exploratory propensity-score matched analysis regarding the type of MI, first diagnosed paroxysmal AF was associated with significantly higher risk of nSTEMI (1.3% vs. 0.7%; HR: 1.89, 95%CI: 1.44–2.46) compared to patients with permanent AF, whereas first diagnosed paroxysmal AF was not associated with higher risk of STEMI (0.5% vs. 0.3%; HR: 1.32, 95%CI: 0.89–1.95).
Table 2.
Risk of myocardial infarction and ischemic stroke in patients with paroxysmal and non-paroxysmal atrial fibrillation
| Paroxysmal AF Events/included patients |
Non-paroxysmal AF Events/included patients |
Hazard ratio (95%CI) |
|
|---|---|---|---|
| Acute myocardial infarction | |||
| All included patients | 260/12 424 | 150/12 424 | 1.65 (1.35–2.01) |
| nSTEMI | 161/12 424 | 81/12 424 | 1.89 (1.44–2.46) |
| STEMI | 60/12 424 | 43/12 424 | 1.32 (0.89–1.95) |
| Excluding those with previous myocardial infarction | 135/11 831 | 75/11 990 | 1.72 (1.29–2.27) |
| Ischemic stroke | |||
| All included patients | 466/12 424 | 409/12 424 | 1.08 (0.95–1.24) |
| Excluding those with previous ischemic stroke | 185/11 687 | 200/11 670 | 0.87 (0.71–1.07) |
| Death | 1287/12 424 | 1625/12 424 | 0.74 (0.69–0.80) |
CI, confidence intervals; nSTEMI, non-ST-elevation myocardial infarction; STEMI, ST-elevation myocardial infarction
In the exploratory propensity-score matched analysis excluding patients with prior MI, 1027 individuals with an MI prior to the diagnosis of AF were excluded. A diagnosis of first-ever acute MI during follow-up was reported in 135 (1.1%) among 11 831 individuals with paroxysmal AF and 75 (0.6%) among 11 990 individuals with non-paroxysmal AF (HR: 1.72, 95%CI: 1.29–2.27) (Table 2).
Type of AF and risk of ischemic stroke and death
After propensity-score matching, 875 (3.5%) were diagnosed with ischemic stroke within 3 years of AF diagnosis, of which 466 (3.8%) occurred in individuals with paroxysmal AF and 409 (3.3%) in individuals with non-paroxysmal AF (HR: 1.08, 95%CI: 0.95–1.24; Supplementary material online, Figure SB) (Table 2).
In the exploratory propensity-score matched analysis excluding patients with prior ischemic stroke, 1491 individuals with an ischemic stroke that occurred before the diagnosis of AF were excluded. A diagnosis of first-ever ischemic stroke during follow-up was reported in 185 (1.6%) among 11 687 individuals with paroxysmal AF and 200 (1.7%) among 11 670 individuals with non-paroxysmal AF (HR: 0.87, 95%CI: 0.71–1.07) (Table 2).
After propensity-score matching 2912 (11.7%) of the patients died during follow-up, of which 1287 (10.4%) among those with paroxysmal AF and 1625 (13.1%) among those with non-paroxysmal AF (HR: 0.74, 95%CI: 0.69–0.80; Supplementary material online, Figure SC) (Table 2).
Discussion
In this propensity-score matched analysis of >24 000 individuals with AF under OAC treatments, who were followed for 3 years, those with first-diagnosed paroxysmal AF had significantly higher risk of acute MI compared to individuals with non-paroxysmal AF. This reduction was associated with a significant reduction of the risk of nSTEMI among patients with first-diagnosed paroxysmal AF compared to those with non-paroxysmal. The risk of ischemic stroke was similar in patients with both paroxysmal and non-paroxysmal-AF.
The association of recently diagnosed paroxysmal AF with higher risk of MI, could potentially be explained through the effect that brief AF episodes exert on the ventricles at a cellular and microcirculatory level. Previous studies in humans have shown that during induced acute AF, there is a beat-to-beat flow variability affecting the diastolic coronary flow and in turn, coronary flow is being reduced compared to patients on sinus rhythm at the same heartrate.13 Recently, a computational analysis based on the simulation of fluid dynamics confirmed that AF and beat-to-beat flow variability significantly reduces coronary flow, particularly in left ventricle’s subendocardial layers.14 Furthermore, during the acute atrial tachyarrhythmia, angiotensin 1-receptor-mediates oxidative stress in the myocardium of the left ventricle and consequent lack of nitric oxide.15,16 In turn, this may lead to impaired microvascular blood flow and ventricular ischaemia.15,16 On the contrary, in patients with chronic AF, effective rate control together with chronic remodelling changes in left atrium may counterbalance the flow variation and oxidative stress, rendering ventricular ischemia less likely.8,15,17 These data may support our results, proposing a potential mechanistical association between first-diagnosed paroxysmal AF and MI, given that patients with paroxysmal AF had higher risk of nSTEMI, rather STEMI. Despite the potential causative and mechanistic associations of first-diagnosed paroxysmal AF on MI, the nature of this association is rather speculative.
The findings of this analysis shed further light on recent contradictory reports about associations between type of AF and the risk of MI, and provide evidence to support findings of recent exploratory analyses of randomized controlled trials suggesting the risk of MI in individuals with AF is associated with the type of AF.8,9,11 In an ENTRUST-AF-PCI subgroup analysis, following successful percutaneous coronary intervention, individuals with paroxysmal AF had high risk of acute MI compared to non-paroxysmal AF (4.87% vs. 2.01%, respectively).8 In an ENSURE-AF subgroup analysis, following electrical cardioversion, MI was more frequent in individuals with paroxysmal vs. persistent AF (1.0% vs. 0.1%, respectively).9 In an EAST-AFNET 4 analysis, the incidence rates of hospitalization for an acute coronary syndrome were highest among individuals with the first AF episode compared to those with paroxysmal and persistent AF (1.50 compared to 0.64 and 0.50, respectively, P for interaction = 0.032).11 Collectively, the results of the present large-scale real-world analysis confirm the hypothesis generated by previous post-hoc exploratory analyses of trials in individuals with AF that paroxysmal AF, and more specifically recently diagnosed paroxysmal AF, may exert an additional risk factor for MI, even in the absence of previous MI.
The association between paroxysmal AF and the risk of MI could have implications for cardiovascular prevention strategies in individuals with paroxysmal AF, especially given that a significant proportion of this population has underlying coronary artery disease. A recent retrospective study showed that coronary CT angiography identified underlying coronary artery disease in 39% of individuals with paroxysmal of first-diagnosed AF.18 In this context, coronary CT angiography is integrated in the diagnostic work-up of individuals with paroxysmal or first-diagnosed AF.18 In addition, intensive strategies aiming to more efficient control of the atherosclerotic risk factors beyond OAC treatment, which could include lifestyle changes like smoking cessation, exercise and physical activity interventions, adoption of Mediterranean diet, and better weight control, as well as pharmaceutical interventions to control related comorbidities such as diabetes mellitus and dyslipidemia more efficiently. Accordingly, several retrospective studies showed that statin treatment in individuals with AF is associated with reduced risk for cardiovascular events.19–21
The present study shows that the risk of ischemic stroke was similar among OAC-treated patients with paroxysmal and non-paroxysmal AF, while patients with non-paroxysmal AF had higher risk of mortality during follow-up. Consistent with our results, a recent prospective Swiss study including 3843 AF patients showed that persistent and permanent AF were not associated with higher risk of ischemic stroke or systemic embolism.10 On the contrary, several observational studies as well as subgroup analyses of randomized controlled trials showed that individuals with non-paroxysmal AF have higher risk of ischemic stroke compared to paroxysmal.22,23 Whether non-paroxysmal and chronic AF is associated with higher risk of future thromboembolic events compared to paroxysmal AF, is still unclear although the inherent risk may be driven by the presence of associated stroke risk factors.24 Of note, the risk of mortality was lower among patients with paroxysmal AF compared to those with non-paroxysmal. This may be potentially attributed to other confounding factors and concomitant diseases which may have risen during follow-up, such as heart failure or renal disease. In the EURObservational Research Programme–Atrial Fibrillation, which included patients with first-ever diagnosed AF, paroxysmal AF was associated with lower risk of mortality compared to non-paroxysmal, while independent predictors of death were heart failure and renal disease.25 In a Danish national cohort, HF diagnosis prior to AF was associated higher absolute risk of death compared to both AF before HF and AF diagnosed concurrently with HF.26
Strengths and limitations
The main strengths of this study are the large number of individuals included in the analysis, the follow-up period of three years, and the use of propensity score matching to control for clinically and prognostically relevant factors and minimize the risk of bias from confounding. Nonetheless, several limitations of this analysis are noteworthy. HCO EMR data are subject to entry errors and data gaps, and some health conditions may be underreported and represent missing data which may affect the results. However, the large number of included patients and data may lessen the effect of entry or capture errors in the outcomes of interest. Also, outcomes which occurred outside the TriNetX network may have not been well captured. Specifically in this study, there were no available data on how paroxysmal AF was diagnosed and whether this was through 24 h Holter or implantable loop recorders, while we couldn’t export specific data regarding the timing of the MI in relation to AF paroxysm. Additionally, there were no available data on AF paroxysms after the diagnosis of paroxysmal AF and we were not able to correlate potential future AF paroxysms with the occurrence of MI or stroke. Regarding the MI subtypes, in the present study only nSTEMI and STEMI were investigated, while the subtype diagnosis of unknown MI, was not investigated and may have affected the results in the exploratory analysis. Moreover, residual confounding may have influenced the results, including socioeconomic status, risk factor control, anticoagulation after control, use of AF interventions, and lifestyle factors27–29 or the CHA2DS2-VASc which are not available in EMR data. Also, as the data included in this cohort originated predominantly from US HCOs, which may not be representative of the wider global population, therefore the generalizability of these results needs to be explored and confirmed beyond this cohort. Finally, we were unable to take in account the effect of ablation or pharmacological rhythm control after the diagnosis since patients could only be matched based on the characteristics before the diagnosis of AF.
In conclusion, this large real-world propensity-score matched analysis shows that among individuals treated with OACs, those with first-diagnosed paroxysmal AF have higher risk of acute MI, a finding which was driven by a significant increase in nSTEMI, compared to individuals with non-paroxysmal AF, while the risk of ischemic stroke was similar in patients with paroxysmal and non-paroxysmal AF. Paroxysmal AF was associated with lower risk of death compared to non-paroxysmal AF. The presence of first-diagnosed paroxysmal AF may serve as cardiovascular risk factor to further investigate and stratify patients potentially at higher risk of acute coronary disease. Whether the type of AF has a causative association with either acute MI or ischemic stroke, needs to be investigated in dedicated prospective randomized trials.
Supplementary Material
Acknowledgements
G.N. contributed to study concept, study design, data acquisition, statistical analysis and interpretation, preparation of article, and study supervision. D.S. contributed to study design, data acquisition, statistical analysis and interpretation, preparation of article, and critical revision of the final draft. B.J.R.B. contributed critical revision of article. S.L.H. contributed to critical revision of article. A.A.R. contributed to critical revision of article. P.A. contributed to critical revision of article. G.Y.H.L. contributed to study concept and critical revision of article.
Contributor Information
George Ntaios, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Department of Internal Medicine, Faculty of Medicine, School of Health Sciences, University of Thessaly, Larissa, Greece.
Dimitrios Sagris, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Department of Internal Medicine, Faculty of Medicine, School of Health Sciences, University of Thessaly, Larissa, Greece.
Benjamin J R Buckley, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Liverpool Centre of Cardiovascular Science, Liverpool John Moores University and Liverpool Heart & Chest Hospital NHS Foundation Trust, Liverpool, UK.
Stephanie L Harrison, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Department of Cardiovascular and Metabolic Medicine, University of Liverpool, Liverpool, UK.
Azmil Abdul-Rahim, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Department of Cardiovascular and Metabolic Medicine, University of Liverpool, Liverpool, UK.
Philip Austin, TriNetX LLC, London, UK.
Gregory Y H Lip, Liverpool Centre of Cardiovascular Science, University of Liverpool, Liverpool, UK; Department of Cardiovascular and Metabolic Medicine, University of Liverpool, Liverpool, UK; Liverpool Centre of Cardiovascular Science, Liverpool John Moores University and Liverpool Heart & Chest Hospital NHS Foundation Trust, Liverpool, UK; Department of Clinical Medicine, Aalborg University, Aalborg, Denmark.
Supplementary material
Supplementary material is available at Europace online.
Funding
TriNetX LLC funded the acquisition of the data used through use of the network database. D.S.’s research fellowship at the Liverpool Centre of Cardiovascular Science, supported by the European Society of Cardiology Council on Stroke.
Conflict of interest
G.N. reported receiving speaker honoraria from Abbott, Amgen, Boehringer Ingelheim, Elpen, and Sanofi paid to the University of Thessaly. D.S. reports receiving research support by the ESC council on Stroke. B.J.R.B. has received investigator-initiated research funding from BMS/Pfizer. S.L.H. has received a grant from Bristol-Myers Squibb (BMS) outside of the submitted work. A.A.R. reports no conflict of interests. P.A. is an employee of TriNetX. G.Y.H.L. reports being a consultant and speaker for BMS/Pfizer, Boehringer Ingelheim and Daiichi-Sankyo. No fees are received personally.
Data availability
The data underlying this article were accessed from TriNetX (https://live.trinetx.com). To gain access to the data in the TriNetX research network, a request can be made to TriNetX (https://live.trinetx.com), but costs may be incurred, a data sharing agreement would be necessary, and no patient identifiable information can be obtained.
References
- 1. Potpara TS, Lip GYH, Blomstrom-Lundqvist C, Boriani G, Van Gelder IC, Heidbuchel Het al. The 4S-AF scheme (stroke risk; symptoms; severity of burden; substrate): a novel approach to in-depth characterization (rather than classification) of atrial fibrillation. Thromb Haemost 2021;121:270–8. [DOI] [PubMed] [Google Scholar]
- 2. Hindricks G, Potpara T, Dagres N, Arbelo E, Bax JJ, Blomström-Lundqvist Cet al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J 2021;42:373–498. [DOI] [PubMed] [Google Scholar]
- 3. Chao TF, Joung B, Takahashi Y, Lim TW, Choi EK, Chan YHet al. 2021 Focused update consensus guidelines of the Asia Pacific Heart Rhythm Society on stroke prevention in atrial fibrillation: executive summary. Thromb Haemost 2022;122:20–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Stevens D, Harrison SL, Kolamunnage-Dona R, Lip GYH, Lane DA. The atrial fibrillation better care pathway for managing atrial fibrillation: a review. Europace 2021;23:1511–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Ding WY, Lane DA, Gupta D, Huisman MV, Lip GYH; GLORIA-AF Investigators . Incidence and risk factors for residual adverse events despite anticoagulation in atrial fibrillation: results from phase II/III of the GLORIA-AF registry. J Am Heart Assoc 2022;11:e026410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Buckley BJR, Harrison SL, Hill A, Underhill P, Lane DA, Lip GYH. Stroke-heart syndrome: incidence and clinical outcomes of cardiac complications following stroke. Stroke 2022;53:1759–63. [DOI] [PubMed] [Google Scholar]
- 7. Romiti GF, Pastori D, Rivera-Caravaca JM, Ding WY, Gue YX, Menichelli Det al. Adherence to the ‘Atrial Fibrillation Better Care’ pathway in patients with atrial fibrillation: impact on clinical outcomes—a systematic review and meta-analysis of 285,000 patients. Thromb Haemost 2022;122:406–14. [DOI] [PubMed] [Google Scholar]
- 8. Goette A, Eckardt L, Valgimigli M, Lewalter T, Laeis P, Reimitz P-Eet al. Clinical risk predictors in atrial fibrillation patients following successful coronary stenting: ENTRUST-AF PCI sub-analysis. Clin Res Cardiol 2021;110:831–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Goette A, Lip GYH, Jin J, Heidbuchel H, Cohen A-A, Ezekowitz Met al. Differences in thromboembolic complications between paroxysmal and persistent atrial fibrillation patients following electrical cardioversion (from the ENSURE-AF study). Am J Cardiol 2020;131:27–32. [DOI] [PubMed] [Google Scholar]
- 10. Blum S, Aeschbacher S, Coslovsky M, Meyre PB, Reddiess P, Ammann Pet al. Long-term risk of adverse outcomes according to atrial fibrillation type. Sci Rep 2022;12:2208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Goette A, Borof K, Breithardt G, Camm AJ, Crijns HJGM, Kuck K-Het al. Presenting pattern of atrial fibrillation and outcomes of early rhythm control therapy. J Am Coll Cardiol 2022;80:283–95. [DOI] [PubMed] [Google Scholar]
- 12. Topaloglu U, Palchuk MB. Using a federated network of real-world data to optimize clinical trials operations. JCO Clin Cancer Inform 2018;2:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Kochiadakis GE, Skalidis EI, Kalebubas MD, Igoumenidis NE, Chrysostomakis SI, Kanoupakis EMet al. Effect of acute atrial fibrillation on phasic coronary blood flow pattern and flow reserve in humans. Eur Heart J 2002;23:734–41. [DOI] [PubMed] [Google Scholar]
- 14. Saglietto A, Fois M, Ridolfi L, De Ferrari GM, Anselmino M, Scarsoglio S. A computational analysis of atrial fibrillation effects on coronary perfusion across the different myocardial layers. Sci Rep 2022;12:841. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Goette A, Kalman JM, Aguinaga L, Akar J, Cabrera JA, Chen SAet al. EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: definition, characterization, and clinical implication. Europace 2016;18:1455–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Goette A, Bukowska A, Dobrev D, Pfeiffenberger J, Morawietz H, Strugala Det al. Acute atrial tachyarrhythmia induces angiotensin II type 1 receptor-mediated oxidative stress and microvascular flow abnormalities in the ventricles. Eur Heart J 2009;30:1411–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Reilly SN, Jayaram R, Nahar K, Antoniades C, Verheule S, Channon KMet al. Atrial sources of reactive oxygen species vary with the duration and substrate of atrial fibrillation: implications for the antiarrhythmic effect of statins. Circulation 2011;124:1107–17. [DOI] [PubMed] [Google Scholar]
- 18. Rottlander D, Saal M, Degen H, Godde M, Horlitz M, Haude M. Diagnostic role of coronary CT angiography in paroxysmal or first diagnosed atrial fibrillation. Open Heart 2021;8:e001638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ntaios G, Papavasileiou V, Makaritsis K, Milionis H, Manios E, Michel Pet al. Statin treatment is associated with improved prognosis in patients with AF-related stroke. Int J Cardiol 2014;177:129–33. [DOI] [PubMed] [Google Scholar]
- 20. Choi JY, Seo WK, Kang SH, Jung JM, Cho KH, Yu Set al. Statins improve survival in patients with cardioembolic stroke. Stroke 2014;45:1849–52. [DOI] [PubMed] [Google Scholar]
- 21. Pastori D, Baratta F, Di Rocco A, Farcomeni A, Del Ben M, Angelico Fet al. Statin use and mortality in atrial fibrillation: a systematic review and meta-analysis of 100,287 patients. Pharmacol Res 2021;165:105418. [DOI] [PubMed] [Google Scholar]
- 22. Ganesan AN, Chew DP, Hartshorne T, Selvanayagam JB, Aylward PE, Sanders Pet al. The impact of atrial fibrillation type on the risk of thromboembolism, mortality, and bleeding: a systematic review and meta-analysis. Eur Heart J 2016;37:1591–602. [DOI] [PubMed] [Google Scholar]
- 23. Ntaios G, Vemmou A, Koroboki E, Savvari P, Makaritsis K, Saliaris Met al. The type of atrial fibrillation is associated with long-term outcome in patients with acute ischemic stroke. Int J Cardiol 2013;167:1519–23. [DOI] [PubMed] [Google Scholar]
- 24. Soliman EZ, Safford MM, Muntner P, Khodneva Y, Dawood FZ, Zakai NAet al. Atrial fibrillation and the risk of myocardial infarction. JAMA Intern Med 2014;174:107–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Boriani G, Laroche C, Diemberger I, Fantecchi E, Popescu MI, Rasmussen LHet al. ‘Real-world’ management and outcomes of patients with paroxysmal vs. non-paroxysmal atrial fibrillation in Europe: the EURObservational Research Programme–Atrial Fibrillation (EORP-AF) General Pilot Registry. Europace 2016;18:648–57. [DOI] [PubMed] [Google Scholar]
- 26. Pallisgaard J, Greve AM, Lock-Hansen M, Thune JJ, Fosboel EL, Devereux RBet al. Atrial fibrillation onset before heart failure or vice versa: what is worst? A nationwide register study. Europace 2022;25:283–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Lim J, Lee SR, Choi EK, Han K-D, Jung J-H, Ahn H-Jet al. Exercise and the risk of dementia in patients with newly diagnosed atrial fibrillation: a nationwide population-based study. J Clin Med 2021;10:3126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Park SH, Lee SR, Choi EK, Lee H, Chung Jewook, Choi Jet al. Low risk of dementia in patients with newly diagnosed atrial fibrillation and a clustering of healthy lifestyle behaviors: a nationwide population-based cohort study. J Am Heart Assoc 2022;11:e023739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kim D, Yang P-S, You SC, Sung J-H, Jang E, Yu HTet al. Association of rhythm control with incident dementia among patients with atrial fibrillation: a nationwide population-based cohort study. Age Ageing 2022;51:afab248. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
TriNetx is a research network utilized for several scientific purposes, compliant with the Health Insurance Portability and Accountability Act and the US federal law which protects the privacy and security of healthcare data, including de-identified data as per the de-identification standard of the HIPAA Privacy Rule (https://trinetx.com/real-world-resources/publications/). To comply with legal frameworks and ethical guidelines guarding against data re-identification, the identity of participating Healthcare Organizations (HCOs) and their individual contribution to each dataset are not disclosed. A typical HCO is a large academic health center with data coming from majority of its affiliates. A single HCO frequently has more than one facility, including main and satellite hospitals. The data are stored on the TriNetX database via a physical server at the institution’s data centre or a virtual hosted appliance. The TriNetX platform comprises of a series of these appliances connected into a federated network. This network can broadcast queries to each appliance. The network members are mainly in the United States, with an additional six HCOs in other countries that include the United Kingdom, Germany, Italy, Israel, and Singapore.12
The TriNetX platform only uses aggregated counts and statistical summaries of de-identified information. No protected health information or personal data is made available to the users of the platform. Studies using the TriNetX health research network do not need ethical approval as no patient identifiable information is received.
To gain access to the data in the TriNetX research network, requests are directed to TriNetX and a data sharing agreement is required.
The data underlying this article were accessed from TriNetX (https://live.trinetx.com). To gain access to the data in the TriNetX research network, a request can be made to TriNetX (https://live.trinetx.com), but costs may be incurred, a data sharing agreement would be necessary, and no patient identifiable information can be obtained.