Predictors of Stroke in Patients With Atrial Fibrillation: A Retrospective Study.

Background Stroke remains one of the most serious complications of atrial fibrillation (AF) and is associated with substantial long-term disability and mortality worldwide. Objective The objective of this study is to determine the frequency and independent predictors of ischemic stroke among patients with AF and assess the associations of demographic, clinical, and echocardiographic factors with stroke occurrence. Methods This retrospective study was conducted at Sir Ganga Ram Hospital, Lahore, Pakistan, from October 2022 to October 2025 and included 225 patients with non-valvular AF. Medical records were reviewed to collect demographic characteristics, comorbidities, laboratory and echocardiographic findings, and anticoagulation status. Stroke events included both prior history and incident events confirmed by CT or MRI during the study period. "Inadequate anticoagulation" was defined as subtherapeutic INR (<2.0) for warfarin users or documented missed/discontinued doses for direct oral anticoagulant users. Multivariable logistic regression was performed to identify independent predictors of stroke, reporting adjusted odds ratios (aORs), 95% confidence intervals (CIs), and p-values. Results Stroke occurred in 78 patients (34.7%). Multivariable analysis identified several independent predictors of stroke. Patients aged 65 years or older had more than twice the odds of stroke (aOR 2.41; 95% CI 1.32-4.39; p = 0.004). Hypertension was also associated with increased risk (aOR 1.98; 95% CI 1.05-3.72; p = 0.03). Echocardiographic findings such as left atrial enlargement (aOR 2.89; 95% CI 1.56-5.36; p = 0.001) and the presence of left atrial appendage (LAA) thrombus or spontaneous echo contrast (SEC) (aOR 3.72; 95% CI 1.72-8.03; p < 0.001) significantly predicted stroke. Inadequate anticoagulation conferred a threefold higher risk (aOR 3.15; 95% CI 1.72-5.77; p < 0.001), and smoking was also an independent predictor (aOR 1.84; 95% CI 1.05-3.23; p = 0.03). Other variables, including diabetes mellitus, dyslipidemia, and reduced left ventricular ejection fraction, were not independently associated with stroke after adjustment. Conclusion In this cohort, stroke in AF patients was strongly associated with older age, uncontrolled hypertension, left atrial enlargement, LAA thrombus/SEC, inadequate anticoagulation, and smoking. Early identification of high-risk individuals and optimization of anticoagulation therapy are critical for reducing thromboembolic events. Prospective studies are warranted to validate these findings and guide risk stratification in broader populations.