Modified Taiwan Atrial Fibrillation Score for the Prediction of Incident Atrial Fibrillation

The optimal antithrombotic strategy for post-stroke patients with atrial fibrillation and extracranial artery stenosis-a nationwide cohort study

BACKGROUND

In post-stroke atrial fibrillation (AF) patients who have indications for both oral anticoagulant (OAC) and antiplatelet agent (AP), e.g., those with carotid artery stenosis, there is debate over the best antithrombotic strategy. We aimed to compare the risks of ischemic stroke, composite of ischemic stroke/major bleeding and composite of ischemic stroke/intracranial hemorrhage (ICH) between different antithrombotic strategies.

METHODS

This study included post-stroke AF patients with and without extracranial artery stenosis (ECAS) (n = 6390 and 28,093, respectively) identified from the Taiwan National Health Insurance Research Database. Risks of clinical outcomes and net clinical benefit (NCB) with different antithrombotic strategies were compared to AP alone.

RESULTS

The risk of recurrent ischemic stroke was higher for patients with ECAS than those without (12.72%/yr versus 10.60/yr; adjusted hazard ratio [aHR] 1.104, 95% confidence interval [CI] 1.052-1.158, p < 0.001). For patients with ECAS, when compared to AP only, non-vitamin K antagonist oral anticoagulant (NOAC) monotherapy was associated with lower risks for ischaemic stroke (aHR 0.551, 95% CI 0.454-0.669), the composite of ischaemic stroke/major bleeding (aHR 0.626, 95% CI 0.529-0.741) and the composite of ischaemic stroke/ICH (aHR 0.577, 95% CI 0.478-0.697), with non-significant difference for major bleeding and ICH. When compared to AP only, warfarin monotherapy was associated with higher risks of major bleeding (aHR 1.521, 95% CI 1.231-1.880), ICH (aHR 2.045, 95% CI 1.329-3.148), and the composite of ischaemic stroke and major bleeding. With combination of AP plus warfarin, there was an increase in ischaemic stroke, major bleeding, and the composite outcomes, when compared to AP only. NOAC monotherapy was the only approach associated with a positive NCB, while all other options (warfarin, combination of AP-OAC) were associated with negative NCB.

CONCLUSIONS

For post-stroke AF patients with ECAS, NOAC monotherapy was associated with lower risks of adverse outcomes and a positive NCB. Combination of AP with NOAC or warfarin did not offer any benefit, but more bleeding especially with AP-warfarin combination therapy.

Risk of Incident Atrial Fibrillation and Subsequent Use of Oral Anticoagulants in Patients with Dementia

BACKGROUND

 Dementia and atrial fibrillation (AF) have many shared risk factors. Besides, patients with dementia are under-represented in randomized trials, and even if AF is present, oral anticoagulants (OACs) are not prescribed frequently. This study aimed to report the incidence of newly diagnosed AF in dementia patients, and the impacts of use of vitamin K antagonist (VKA; e.g., warfarin) and non-VKA OAC (NOACs) on stroke and bleeding outcomes.

METHODS

 Our study utilized the Taiwan National Health Insurance Research Database. A total of 554,074 patients with dementia were compared with 554,074 age- and sex-matched patients without dementia regarding the risk of incident AF. Among patients with dementia who experienced incident AF, the risks of clinical events of patients treated with warfarin or NOACs were compared with those without OACs (reference group).

RESULTS

 The risk of incident AF was greater for patients with dementia compared with those without (adjusted hazard ratio [aHR]: 1.054; 95% confidence interval [CI]: 1.040-1.068 for all types of dementia, aHR: 1.035; 95% CI: 1.020-1.051 for presenile/senile dementia, and aHR: 1.125; 95% CI: 1.091-1.159 for vascular dementia). Among patients with dementia and experienced incident AF, warfarin use was associated with a higher risk of ischemic stroke (aHR: 1.290; 95% CI: 1.156-1.440), intracranial hemorrhage (ICH; aHR: 1.678; 95% CI: 1.346-2.090), and major bleeding (aHR: 1.192; 95% CI: 1.073-1.323) compared with non-OACs. NOAC use was associated with a lower risk of ischemic stroke (aHR: 0.421; 95% CI: 0.352-0.503) and composite risk of ischemic stroke or major bleeding (aHR: 0.544; 95% CI: 0.487-0.608) compared with non-OACs. These results were consistent among the patients after the propensity matching.

CONCLUSION

 In this large nationwide cohort, the risk of newly diagnosed AF was higher in patients with dementia (all dementia, presenile/senile dementia, and vascular dementia) compared with those without dementia. For patients with dementia who experienced incident AF, NOAC use was associated with a better clinical outcome compared with non-OAC. Patients with dementia require a holistic approach to their care and management, including the use of NOACs to reduce the risks of clinical events.

Automated risk assessment of newly detected atrial fibrillation poststroke from electronic health record data using machine learning and natural language processing

Background

Timely detection of atrial fibrillation (AF) after stroke is highly clinically relevant, aiding decisions on the optimal strategies for secondary prevention of stroke. In the context of limited medical resources, it is crucial to set the right priorities of extended heart rhythm monitoring by stratifying patients into different risk groups likely to have newly detected AF (NDAF). This study aimed to develop an electronic health record (EHR)-based machine learning model to assess the risk of NDAF in an early stage after stroke.

Methods

Linked data between a hospital stroke registry and a deidentified research-based database including EHRs and administrative claims data was used. Demographic features, physiological measurements, routine laboratory results, and clinical free text were extracted from EHRs. The extreme gradient boosting algorithm was used to build the prediction model. The prediction performance was evaluated by the C-index and was compared to that of the AS5F and CHASE-LESS scores.

Results

The study population consisted of a training set of 4,064 and a temporal test set of 1,492 patients. During a median follow-up of 10.2 months, the incidence rate of NDAF was 87.0 per 1,000 person-year in the test set. On the test set, the model based on both structured and unstructured data achieved a C-index of 0.840, which was significantly higher than those of the AS5F (0.779, p = 0.023) and CHASE-LESS (0.768, p = 0.005) scores.

Conclusions

It is feasible to build a machine learning model to assess the risk of NDAF based on EHR data available at the time of hospital admission. Inclusion of information derived from clinical free text can significantly improve the model performance and may outperform risk scores developed using traditional statistical methods. Further studies are needed to assess the clinical usefulness of the prediction model.