Prognostic value of routine blood tests along with clinical risk factors in predicting ischemic stroke in non-valvular atrial fibrillation: a prospective cohort study

Background In patients with atrial fibrillation (AF), most biomarkers are still of limited use due to cost-effectiveness and complexity in clinical practice. Hypotheses Biomarkers from routine blood tests improve the current risk stratification in AF patients. Methods This prospective study enrolled 600 patients diagnosed with non-valvular AF, of whom 537 were analyzed. Platelet count; platelet distribution width (PDW); red cell distribution width (RDW); and creatinine, D-dimer, and troponin I levels were measured at enrollment. Results During the mean follow-up period (2.2 ± 0.6 years), 1.9% patients developed ischemic stroke. According to the optimal cutoff of each biomarker, the risk of ischemic stroke was higher in patients with RDW ≥ 13.5%, creatinine ≥ 1.11 mg/dL, or PDW ≥ 13.2% (significant biomarkers; P value: < 0.01, 0.04, or 0.07, respectively). These 3 significant biomarkers had higher information gain than clinical risk factors in predicting ischemic stroke. The cumulative incidence of ischemic stroke was 1.2%, 1.1%, 8.4%, and 40.0% in patients with 0, 1, 2, and 3 significant biomarkers, respectively (P-for-trend < 0.001). Patients with ≥ 2 significant biomarkers had a significantly higher risk of ischemic stroke than those with < 2 significant biomarkers (adjusted hazard ratio 11.5, 95% confidence interval 3.3–40.2, P < 0.001). The predictability for ischemic stroke was significantly improved when ≥ 2 significant biomarkers were added to the CHA2DS2–VASc score (area under the curve 0.790 vs. 0.620, P = 0.043). Conclusion Routine blood tests can provide better risk stratification of AF along with clinical risk factors.