ATRIA Bleeding Risk Score

• The ATRIA risk scoring scheme, in combination with other risk stratification scores and physician gestalt, can be used to help guide the decision for warfarin therapy in patients in need of anticoagulation.
• The ATRIA risk scoring scheme may be used as one potential tool to stratify the risk of major hemorrhage in patients in whom warfarin anticoagulation is being considered.

The ATRIA study sought to develop a risk stratification score to predict warfarin-associated hemorrhage.

• Included patients from the Kaiser Permanente of Northern California hospital system with non-valvular, non-transient atrial fibrillation via their ICD-9 codes and then followed up for a period of six to seven years.
• Patients randomly divided into split-sample “derivation” and “validation” cohorts, in a 2:1 ratio, respectively.
• The study sought to identify risk factors for patients with major hemorrhage, defined as fatal, requiring transfusion of ≥ 2 units pRBCs, or hemorrhage into critical anatomic site (intracranial, retroperitoneal, etc).
• Five variables were derived after statistical analysis and then weighed based on degree of effect.
• Scoring system consisted of five criteria, each of which was worth 1, 2, or 3 points.
• Study used ≤ 3 points to define low probability group (<1%) for major hemorrhage, 4 points for immediate (2.6%) and 5-10 points for high (5.8%).

Points to keep in mind:

• Patients at risk for major hemorrhage, e.g. those with past intracranial bleeds, were likely already pre-excluded by clinicians and thus not included in the study
• The derivation and validation cohorts were taken from the same patient population with excellent follow-up
• Patients’ INRs and potential for compliance were not taken into consideration
• The study did not compare the ATRIA score to the HAS-BLED score, which has been incorporated into European and Canadian atrial fibrillation guidelines
• Subsequent studies comparing the performance of the ATRIA and HAS-BLED bleeding risk-prediction scores showed that HAS-BLED had better prediction accuracy than ATRIA.

Considers major and minor risk of hemorrhage and bleeding in patients prior to beginning warfarin therapy