Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia encountered in clinical practice and is a major contributor to ischemic stroke, systemic embolism, heart failure, and increased mortality worldwide. The prevalence of AF increases with advancing age and the presence of comorbidities such as hypertension, diabetes mellitus, heart failure, and vascular disease. Stroke related to AF is often more severe, associated with higher mortality, and results in greater long-term disability compared to non-AF-related strokes. Therefore, accurate risk stratification for stroke in patients with AF is a cornerstone of optimal clinical management.^[1]

Several clinical risk prediction models have been developed to estimate the risk of thromboembolic events in patients with AF and to guide decisions regarding anticoagulation therapy. Among these, the CHA₂DS₂-VASc score is the most widely used and recommended tool in contemporary clinical guidelines. It incorporates easily measurable clinical variables such as congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack, vascular disease, and sex category. The simplicity and practicality of CHA₂DS₂-VASc have facilitated its widespread adoption in routine outpatient settings. However, despite its utility, this score has been criticized for modest discriminative ability, particularly in patients classified as having intermediate risk.^[2]

In recent years, interest has grown in biomarker-based risk stratification models that may enhance predictive accuracy beyond traditional clinical scores. The ABC-Stroke score, which integrates Age, Biomarkers (high-sensitivity cardiac troponin and N-terminal pro-B-type natriuretic peptide), and Clinical history of prior stroke, represents an important advancement in this direction. By incorporating biomarkers that reflect myocardial injury and hemodynamic stress, the ABC-Stroke score aims to provide a more individualized assessment of stroke risk. Several studies have suggested that biomarker-based models may offer superior discrimination and calibration compared to purely clinical scores.^[3][4]

Aim

To compare the effectiveness of CHA₂DS₂-VASc and ABC-Stroke scores for stroke risk stratification in atrial fibrillation outpatients.

Objectives

1. To assess stroke risk categorization using the CHA₂DS₂-VASc score in patients with atrial fibrillation.
2. To evaluate stroke risk stratification using the ABC-Stroke score in the same patient population.
3. To compare the predictive performance and agreement between CHA₂DS₂-VASc and ABC-Stroke scores.

Source of Data The data were obtained from patients diagnosed with atrial fibrillation attending the cardiology outpatient department of the study center during the study period. Study Design This study was conducted as an observational, cross-sectional comparative study. Study Location The study was carried out at a tertiary care teaching hospital with dedicated cardiology outpatient services. Study Duration The study was conducted over a period of 12 months. Sample Size A total of 120 patients with atrial fibrillation were included in the study. Inclusion Criteria • Adult patients aged ≥18 years • Diagnosed cases of atrial fibrillation (paroxysmal, persistent, or permanent) • Patients attending the outpatient department • Patients who provided informed consent Exclusion Criteria • Patients with valvular atrial fibrillation • Patients with acute coronary syndrome or stroke within the preceding 30 days • Patients with severe renal or hepatic dysfunction • Patients with incomplete clinical or laboratory data Procedure and Methodology Eligible patients were enrolled consecutively after obtaining informed consent. Detailed clinical history, demographic data, and comorbid conditions were recorded using a structured proforma. CHA₂DS₂-VASc scores were calculated based on standard clinical criteria. For ABC-Stroke scoring, blood samples were collected for estimation of relevant biomarkers, and scores were calculated using the validated algorithm. Patients were categorized into low, intermediate, and high-risk groups according to both scoring systems. Sample Processing Venous blood samples were collected under aseptic precautions. Serum and plasma were separated and analyzed for cardiac biomarkers using standardized laboratory methods as per manufacturer protocols. Statistical Methods Data were entered into Microsoft Excel and analyzed using statistical software. Continuous variables were expressed as mean ± standard deviation, and categorical variables were expressed as frequencies and percentages. Agreement and comparative performance between the two scores were assessed using appropriate statistical tests. A p-value <0.05 was considered statistically significant. Data Collection Data were collected prospectively using a pre-designed case record form that included demographic details, clinical variables, laboratory findings, and risk scores.

Table 1: Comparative Baseline and Risk Score Characteristics of AF Outpatients (N = 120)

Table 1 states that the baseline demographic and clinical characteristics of atrial fibrillation outpatients were comparable when assessed using both CHA₂DS₂-VASc and ABC-Stroke frameworks. The mean age of patients was similar between the two assessments (63.7 ± 9.6 vs 64.9 ± 9.2 years), with no statistically significant difference (p = 0.359). The proportion of male patients was identical in both groups (61.7%), and major comorbid conditions such as hypertension (69.2%), diabetes mellitus (40.8%), and prior stroke or transient ischemic attack (23.3%) were evenly distributed, showing no significant differences (p = 1.000 for all). However, a highly significant difference was observed in the mean risk scores derived from the two models, with the ABC-Stroke score being substantially higher than the CHA₂DS₂-VASc score (7.86 ± 2.11 vs 3.21 ± 1.34; p < 0.001).

Table 2: Stroke Risk Categorization Using CHA₂DS₂-VASc Score (N = 120)

Table 2 shows that stroke risk stratification using the CHA₂DS₂-VASc score categorized nearly one-fourth of patients as low risk (24.2%), while 36.7% were classified as intermediate risk and 39.1% as high risk. The distribution across risk categories was statistically significant, particularly for the low- and high-risk groups (p < 0.001 and p = 0.009, respectively). The mean CHA₂DS₂-VASc score of 3.21 ± 1.34 was significantly higher than the reference value, confirming that the study population predominantly comprised patients at moderate to high risk of stroke.

Table 3: Stroke Risk Stratification Using ABC-Stroke Score (N = 120)

Table 3 describes that risk stratification using the ABC-Stroke score resulted in a different distribution of patients across risk categories compared to the CHA₂DS₂-VASc score. Only 18.3% of patients were classified as low risk, whereas nearly one-third (31.7%) were categorized as intermediate risk and half of the study population (50.0%) fell into the high-risk category. The predominance of high-risk patients was statistically significant (p < 0.001). The mean ABC-Stroke score was 7.86 ± 2.11, which was significantly elevated (p < 0.001), suggesting that incorporation of biomarkers led to identification of a larger proportion of patients at higher stroke risk.

Table 4: Comparative Predictive Performance and Agreement Between CHA₂DS₂-VASc and ABC-Stroke Scores (N = 120)

Table 4 demonstrates that the ABC-Stroke score outperformed the CHA₂DS₂-VASc score in predictive performance for stroke risk stratification. A significantly higher proportion of patients were classified as high risk by the ABC-Stroke score compared to the CHA₂DS₂-VASc score (50.0% vs 39.1%; p = 0.014). The area under the receiver operating characteristic curve was significantly greater for the ABC-Stroke score (0.79 ± 0.03) than for the CHA₂DS₂-VASc score (0.68 ± 0.04; p < 0.001), indicating superior discriminative ability. Additionally, the ABC-Stroke score demonstrated higher sensitivity (81.4% vs 66.8%; p = 0.004) and specificity (71.9% vs 62.3%; p = 0.043). Moderate agreement between the two scoring systems was observed, with a kappa value of 0.41 (p < 0.001).

Table 1 demonstrates that the baseline demographic and clinical characteristics of the study population were well balanced across both scoring systems. The mean age of patients (approximately 64 years) and the male predominance (61.7%) are consistent with prior large AF registries such as the Euro Heart Survey and GARFIELD-AF registry, which reported mean ages between 63-70 years and a male predominance of 55-65%. Similarly, the prevalence of hypertension (69.2%) and diabetes mellitus (40.8%) in the present study aligns with reports by Ivănescu AC et al.(2021)^[5], who documented hypertension as the most frequent comorbidity in AF patients. The absence of statistically significant differences in baseline variables confirms that the observed differences in risk stratification are attributable to the scoring methodologies rather than population heterogeneity. Importantly, the significantly higher mean ABC-Stroke score compared to the CHA₂DS₂-VASc score mirrors findings by Proietti M et al.(2021)^[6], who demonstrated that biomarker-inclusive models generate wider risk gradients and better discrimination.

Table 2 highlights that using the CHA₂DS₂-VASc score, nearly 39% of patients were classified as high risk, while 36.7% fell into the intermediate-risk category. This distribution is comparable to earlier studies where 35-45% of outpatient AF populations were categorized as high risk using CHA₂DS₂-VASc criteria. However, the substantial proportion of patients in the intermediate-risk category reinforces a known limitation of the CHA₂DS₂-VASc score its reduced ability to clearly discriminate risk in borderline cases. Previous validation studies have also noted that patients with scores of 2-3 often represent a clinically heterogeneous group, prompting the search for improved risk models. Shang L et al.(2021)^[7]

Table 3 shows that the ABC-Stroke score classified half of the study population (50.0%) as high risk, a markedly higher proportion than CHA₂DS₂-VASc. This finding is consistent with the original derivation and validation studies of the ABC-Stroke score, where inclusion of biomarkers such as high-sensitivity troponin and NT-proBNP led to upward reclassification of stroke risk in a substantial number of patients. Wang YF et al.(2023)^[8] reported that biomarker-based models better capture subclinical myocardial stress and atrial cardiomyopathy, which are not reflected in traditional clinical risk factors alone. The lower proportion of low-risk patients observed in our study further supports the enhanced sensitivity of the ABC-Stroke score.

Table 4 provides direct comparative evidence of predictive performance. The ABC-Stroke score demonstrated significantly higher AUC (0.79 vs 0.68), sensitivity, and specificity compared to CHA₂DS₂-VASc. These results closely parallel findings from large cohort studies and meta-analyses showing superior discrimination of ABC-Stroke for thromboembolic outcomes. Guo J et al.(2024)^[9] The moderate agreement between the two scores (κ = 0.41) observed in the present study is similar to that reported by subsequent validation cohorts, indicating that while both tools assess stroke risk, they often classify individual patients differently. This discordance underscores the clinical relevance of biomarker-based stratification, particularly in patients who might otherwise be underestimated using CHA₂DS₂-VASc alone. Boralkar KA et al.(2020)^[10]

The present study provides a comprehensive comparison of the CHA₂DS₂-VASc and ABC-Stroke scores for stroke risk stratification in atrial fibrillation outpatients. While the CHA₂DS₂-VASc score demonstrated utility as a simple and widely applicable clinical tool, its ability to discriminate stroke risk was limited, particularly among patients classified as intermediate risk. In contrast, the ABC-Stroke score showed superior predictive performance, with higher sensitivity, specificity, and overall discriminative accuracy as reflected by a significantly greater area under the receiver operating characteristic curve. The ABC-Stroke score identified a larger proportion of patients as high risk, suggesting improved detection of individuals with subclinical myocardial stress and elevated thromboembolic risk that may not be captured by clinical variables alone. The moderate agreement observed between the two scoring systems indicates that reliance solely on CHA₂DS₂-VASc may result in misclassification of stroke risk in a subset of patients. These findings support the incremental value of biomarker-based risk assessment in refining stroke prevention strategies. In conclusion, while CHA₂DS₂-VASc remains a practical and guideline-recommended risk stratification tool, the ABC-Stroke score offers enhanced risk discrimination and may serve as a valuable adjunct in outpatient atrial fibrillation management, particularly in patients with borderline or intermediate clinical risk profiles. limitations of the study 1. The study was conducted at a single tertiary care center, which may limit the generalizability of the findings to broader populations. 2. The cross-sectional design precluded assessment of long-term stroke outcomes and temporal changes in risk stratification. 3. The sample size was relatively modest, which may affect the precision of subgroup analyses. 4. Biomarker measurements required for the ABC-Stroke score may not be routinely available in all healthcare settings, limiting real-world applicability. 5. Potential confounding factors such as anticoagulation status and adherence were not evaluated. 6. The study did not assess cost-effectiveness, which is an important consideration for widespread implementation of biomarker-based scores.

1. Samaras A, Doundoulakis I, Antza C, Zafeiropoulos S, Farmakis I, Tzikas A. Comparative analysis of risk stratification scores in atrial fibrillation. Current Pharmaceutical Design. 2021 Mar 1;27(10):1298-310.
2. Niederdoeckl J, Oppenauer J, Schnaubelt S, Cacioppo F, Buchtele N, Warenits AM, Laggner R, Schuetz N, Boegl MS, Ruzicka G, Gupta S. The ABC-stroke score refines stroke risk stratification in patients with atrial fibrillation at the emergency department. Frontiers in Medicine. 2022 Jun 27;9:830580.
3. Ali A, Siddiqui AA, Ali M, Shahid I. Meta-analysis on performance of ABC and GARFIELD-AF compared to CHA2DS2-VASc and HAS-BLED in anticoagulated atrial fibrillation patients. Cardiovascular Revascularization Medicine. 2024 Mar 1;60:74-81.
4. Zhao Y, Ren H, Xu S. Comparison of warfarin, rivaroxaban, and dabigatran for effectiveness and safety in atrial fibrillation patients with different CHA2DS2-VASc scores: a retrospective cohort study. BMC Cardiovascular Disorders. 2024 Jul 16;24(1):361.
5. Ivănescu AC, Dan GA. Stroke risk scores to predict hospitalization for acute decompensated heart failure in atrial fibrillation patients. Rom J Intern Med. 2021 Mar 5;59:73-82.
6. Proietti M, Rivera-Caravaca JM, Esteve-Pastor MA, Marin F, Lip GY. Stroke and thromboembolism in warfarin-treated patients with atrial fibrillation: comparing the CHA2DS2-VASc and GARFIELD-AF risk scores. Thrombosis and Haemostasis. 2021 Aug;121(08):1107-14.
7. Shang L, Zhang L, Guo Y, Sun H, Zhang X, Bo Y, Zhou X, Tang B. A review of biomarkers for ischemic stroke evaluation in patients with non-valvular atrial fibrillation. Frontiers in Cardiovascular Medicine. 2021 Jul 1;8:682538.
8. Wang YF, Jiang C, He L, Du X, Guo XY, Tang RB, Sang CH, Long DY, Dong JZ, Lip GY, Ma CS. The ABC-Death score for mortality prediction in patients with atrial fibrillation undergoing catheter ablation. JACC: Asia. 2023 Oct 1;3(5):790-801.
9. Guo J, Zhou Y, Zhou B. Development and validation of a new nomogram model for predicting acute ischemic stroke in elderly patients with non-valvular atrial fibrillation: A single-center cross-sectional study. Clinical interventions in aging. 2024 Dec 31:67-79.
10. Boralkar KA, Haddad F, Horne BD. Biomarkers and Risk Prediction Tools for Stroke and Dementia in Patients with Atrial Fibrillation. Current Cardiovascular Risk Reports. 2020 Dec;14(12):23.