Prediction of bleeding events in patients with non-valvular atrial fibrillation - validation of the VTE-BLEED score

Prediction of bleeding events in patients with non-valvular atrial fibrillation - validation of the VTE-BLEED score

None

Research University of Leiden University Medical Center

Background
Anticoagulation treatment in patients with atrial fibrillation (AF), which is not associated with heart valve abnormalities, is unavoidably associated with an increased risk of bleeding, regardless of anticoagulation type used. In recent years several clinical prediction scores have been developed and validated in large cohorts, and can be used as tools for bleeding risk evaluation in AF patients, namely: HAS-BLED, ORBIT, ATRIA, HEMORR2HAGES and ABC-bleeding scores.
So far, few studies have focused on comparisons and validation of the different bleeding scores in patients treated with direct oral anticoagulants (DOACs). A recent comparison of the predictive performances of these tools in patients from the RE-LY study, in which a comparison between dabigatran and VKA was performed in patients with atrial fibrillation, has showed that all bleeding risk scores identified patient groups with different risks of major bleeding and life-threatening bleeding with only modest and variable discriminative ability.
Recently, the 6-variable VTE-BLEED score has been derived after analysis of consecutive patients with venous thromboembolism included in the two RE-COVER trials, who were randomised to treatment with dabigatran or VKA to compare the efficacy and safety of these regimens. This score proved accurate for the prediction of major bleeding in patients treated with dabigatran both before and after day 30 during a six-month follow up, both in patients on dabigatran and those on warfarin. The VTE-BLEED score encompasses six easy obtainable variables.
Objectives
To test the predictive performance of the VTE-BLEED score for major bleeding in AF patients enrolled in the RE-LY study.
Methods
We will perform a post-hoc analysis of patients included in the RE-LY study. The diagnostic quality of the VTE-BLEED score will be examined using receiver operator characteristic curve analysis (c-statistic) by plotting the true positive rate against the false positive rate at various threshold settings of the VTE BLEED score. The bleeding score will be considered to discriminate well in the derivation set if its c-statistic is significantly greater than 0.5 (lower limit of 95% confidence interval above the value of 0.5). Subsequently, we will evaluate the performance of the VTE-BLEED score in predefined subgroups and clinical scenarios. Finally, we will assess the incremental predictive value of other relevant variables not included in the VTE-BLEED.
Relevance
Because the VTE BLEED score comprises easy obtainable variables and performs very well in databases of patients with VTE, and current bleeding scores have been shown to predict bleeding in AF with only modest and variable discriminative ability, the VTE-BLEED could improve bleeding assessment in AF when a high predictive quality is shown.

[{ "PostingID": 2599, "Title": "BI-1160.26", "Description": "Randomized Evaluation of Long Term Anticoagulant Therapy (RE-LY) With Dabigatran Etexilate

Medicine: dabigatran etexilate, Condition: Atrial Fibrillation, Stroke, Phase: 3, Clinical Study ID: 1160.26, Sponsor: Boehringer Ingelheim" }]

The publication citation will be added after the research is published.