Introduction: Our objective was to evaluate the risk of stroke or vascular event recurrence in the Spanish population according to the Essen risk scale (ERS) and suggest modifications to improve its prediction sensitivity and specificity.
Methods: Patients in the Real Action Guidelines in Stroke Care and Treatment Study (PRACTIC), a prospective registry of stroke patients admitted to 88 Spanish hospitals representing the three level of care, were followed- up for 6 months. Sensitivity and specificity in stroke or vascular event recurrence prediction were determined with the ERS and after modifications of this scale.
Results: Sensitivity and specificity for stroke recurrence prediction were 61.5 % and 45.3 % in high-risk patients (p=0.340), 3.8% and 99.3% in very high risk patients (p=0.017). Prediction, sensitivity and specificity were 67.5 % and 46.2 % (p=0.020) or 5% and 99.6% (p=0.017) for any vascular event, respectively. After risk dichotomization, patients scoring >3 points had sensitivity and specificity of 46.2% and 71% (p=0.009) for stroke recurrence and 46.3% and 71.7% for a new vascular event. These results can be optimized by adding the alcoholism variable to the scale, obtaining a sensitivity and specificity of 51.9% and 67% (p=0.005) or 53.8% and 68% (p=0.0001), respectively. We also reach a higher predictive power adding alcoholism: 61.2 % (52.9 %- 69.5 %) vs 56.7% (47.9%-65.6%) for stroke recurrence and 63.7% (57.0%-70.4%) vs 60.5% (53.5%-67.4%) for new vascular event.
Conclusions: ERS has a low sensitivity and predictive power for stroke or new vascular event in the Spanish population. Dichotomization and adding the variable alcoholism improve its sensitivity and predictive power for stroke recurrence either for stroke or for any vascular recurrence. However it is still necessary to modify the risk scale to improve identification of the higher risk patients.