June 24, 2026American Heart Journal0 citations
View Full PaperDOAC Score Among Patients Receiving Vitamin K Antagonists
Key Result
The DOAC Score outperformed the HAS-BLED score for predicting 1-year major bleeding in AF patients taking VKAs in COMBINE-AF (C-statistic 0.62 vs 0.59) and GARFIELD-AF (0.65 vs 0.62) (both P<0.001).
Key Points
- This research aims to evaluate the effectiveness of the DOAC Score in predicting major bleeding risk among patients taking vitamin K antagonists.
- Data was collected from two sources: COMBINE-AF (clinical trials) and GARFIELD-AF (registry).
- Patients were categorized by their DOAC Score into risk groups and major bleeding rates were compared.
- Discrimination performance of the DOAC Score was assessed using C-statistics and compared with HAS-BLED.
- In COMBINE-AF, the one-year major bleeding rate increased with higher DOAC Score categories, with very high risk score patients experiencing 7.9 events/100 p-y.
- C-statistic for DOAC Score in COMBINE-AF was 0.62, significantly higher than HAS-BLED at 0.59 (P<0.001).
- In GARFIELD-AF, very high risk score patients had a major bleeding rate of 7.6 events/100 p-y, with C-statistic of 0.65, better than HAS-BLED at 0.62 (P<0.001).
Study Design
Type
Cohort (n=49,001)
Multicenter
Yes
Structured PICO
Does the DOAC Score improve risk stratification for major bleeding compared to the HAS-BLED score in patients with atrial fibrillation taking vitamin K antagonists?
P
Population
49,001 patients with atrial fibrillation receiving vitamin K antagonists (28,818 from COMBINE-AF pooled clinical trials and 20,183 from GARFIELD-AF registry).
I
Intervention
Risk stratification using the DOAC Score
C
Comparator
Risk stratification using the HAS-BLED score
O
Outcome
Major bleeding at one-yearsafety
The DOAC Score outperforms the HAS-BLED score in predicting one-year major bleeding risk among patients with atrial fibrillation treated with vitamin K antagonists.
Main Result
Effect estimate: C-statistic
Absolute Event Rate: 0.62% vs 0.59%
p-value: p=<0.001
Abstract
BACKGROUND: The DOAC Score is a bleeding risk score that incorporates ten common clinical variables to risk stratify major bleeding in patients with atrial fibrillation and demonstrated improved risk stratification than HAS-BLED in patients receiving direct acting oral anticoagulants (DOACs). This study evaluates the discriminative performance of the DOAC Score among patients taking vitamin K antagonists (VKAs). METHODS: Data was obtained from COMBINE-AF and GARFIELD-AF. COMBINE-AF included patients with atrial fibrillation randomized to warfarin from four clinical trials: RE-LY, ARISTOTLE, ROCKET-AF, and ENGAGE AF-TIMI 48. GARFIELD-AF included patients with atrial fibrillation prescribed VKAs in a registry. The DOAC Score of each patient was determined, based on commonly obtained clinical variables. Patients were then stratified by DOAC Score clinical risk categories (very low score: 0-3, low score: 4-5, moderate score: 6-7, high score: 8-9, and very high score: 10), and the rate of major bleeding at one-year was compared between groups. Discrimination was assessed using C-statistics and compared with HAS-BLED using DeLong's test. RESULTS: A total of 28,818 patients in COMBINE-AF and 20,183 patients in GARFIELD-AF receiving vitamin K antagonists were included. Of these individuals, 994 (3.4%) in COMBINE-AF and 313 (1.6%) in GARFIELD-AF experienced a major bleeding event at one-year. Patients in higher DOAC Score risk categories experienced greater one-year major bleeding rates in COMBINE-AF, including very low (1.8 events per 100 person-years events/100p-y), low (3.0 events/100 p-y), moderate (4.6 events/100 p-y), high (5.6 events/100 p-y), and very high (7.9 events/100 p-y). Discrimination in COMBINE-AF was moderate and higher than HAS-BLED at one-year (C-statistic: 0.62 vs 0.59, P<0.001). In GARFIELD-AF, higher risk categories also had higher one-year major bleeding rates: very low (0.8 events per 100 person-years events/100 p-y), low (1.5 events/100 p-y), moderate (2.2 events/100 p-y), high (3.2 events/100 p-y), and very high (7.6 events/100 p-y). Discrimination in GARFIELD-AF was moderate and higher than the HAS-BLED score at one-year (C-statistic: 0.65 vs 0.62, P<0.001). CONCLUSION: In patients with atrial fibrillation taking VKAs, the DOAC Score was able to risk stratify patients based on bleeding risk, had moderate discrimination, and out-performed the HAS-BLED score in both a pooled clinical trials cohort and a usual care registry.