Evaluación dinámica del riesgo tromboembólico y hemorrágico en pacientes con fibrilación auricular anticoagulados con acenocumarol

Abstract

Background: The risk of suffering an ischemic stroke and/or hemorrhage in patients with atrial fibrillation (AF) is often assessed at the time of diagnosis or at the start of anticoagulation therapy, and this result is used to predict events that occur many years later. However, these risks are not static. Therefore, in the present study we investigate whether the dynamic changes in thromboembolic and hemorrhagic risk, estimated using the CHA2DS2-VASc and HAS-BLED scores, respectively, modify the prediction of ischemic stroke and major bleeding. Objectives: To investigate the changes of CHA2DS2-VASc and HAS-BLED scores during the follow-up; to analyze whether the acquisition of new risk factors and age modify the risk category initially assigned to each patient; to compare the predictive ability for ischemic stroke/transient ischemic attack (TIA) and major bleeding of the scores calculated at baseline and dynamically, every two years; and to assess the clinical usefulness and net benefit of using the dynamic scores versus baseline scores to predict future adverse events in order to evaluate if dynamic assessment improves the predictive ability of adverse events regarding the baseline assessment. Methods: Observational study, with prospective longitudinal follow-up in a single centre. From May 1, 2007, to December 1, 2007, outpatients aged ≥18 with permanent or paroxysmal AF who were stable on oral anticoagulation therapy with acenocoumarol (INR 2,0-3,0) in the preceding 6 months, were consecutively included. During 6-years of follow-up, all adverse events were recorded. CHA2DS2-VASc and HAS-BLED scores were recalculated every 2-years, and tested for clinical outcomes in periods of 2-years. Results: We included 1361 patients. Compared to the baseline CHA2DS2-VASc, the CHA2DS2-VASc re-calculated at 2-years and at 4-years presented higher predictive ability for stroke/TIA during the period from 2-4 years and 4-6 years respectively, as well as an improvement in sensitivity and reclassification. The re-calculated HAS-BLED at 2-years presented higher predictive ability compared to the baseline for major bleeding during the period from 2-4 years, with significant improvements in terms of sensitivity and reclassification. Although HAS-BLED score re-calculated at 4-years was non-significantly different than HAS-BLED at baseline, a slight enhance in the sensitivity was observed. On the other hand, the decision curve analysis demonstrated that using the dynamic CHA2DS2-VASc and HAS-BLED scores was clinical useful, and provided an overall improvement in the net benefit for the prediction of adverse events and major bleeding, respectively, compared with the baseline estimate of risks. Conclusions: The variables that define ischemic and bleeding risk have a dynamic nature. Therefore, the dynamic assessment of these risks, through the reassessment of CHA2DS2-VASc and HAS-BLED scores, entails significant modifications of total score regarding that calculate at baseline and implies an increase in the punctuation of the corresponding scores, with a change in the initially assigned risk category. This recategorization of risk (ischemic and bleeding), shows a trend towards an increase in the number of patients reclassified within the high-risk group, to the detriment of a decrease in the proportion of low and moderate-risk patients. The dynamic assessment of the CHA2DS2-VASc and HAS-BLED scores represents an improvement in the predictive ability of adverse events regarding baseline scores. Furthermore, the dynamic assessment was associated with a higher net benefit for the prediction of ischemic stroke/TIA and major bleeding, and therefore with increased clinical usefulness than the baseline scores. Thus, in clinical practice, the most recent risk assessment will be the most reliable for estimating the real risk to which patients with AF are exposed.