Major bleeding and intracranial hemorrhage risk prediction in patients with atrial fibrillation: Attention to modifiable bleeding risk factors or use of a bleeding risk stratification score? A nationwide cohort study

Tze Fan Chao, Gregory Y.H. Lip*, Yenn Jiang Lin, Shih Lin Chang, Li Wei Lo, Yu Feng Hu, Ta Chuan Tuan, Jo Nan Liao, Fa Po Chung, Tzeng Ji Chen, Shih Ann Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

68 Scopus citations

Abstract

Background: While modifiable bleeding risks should be addressed in all patients with atrial fibrillation (AF), use of a bleeding risk score enables clinicians to ‘flag up’ those at risk of bleeding for more regular patient contact reviews. We compared a risk assessment strategy for major bleeding and intracranial hemorrhage (ICH) based on modifiable bleeding risk factors (referred to as a ‘MBR factors’ score) against established bleeding risk stratification scores (HEMORR2HAGES, HAS-BLED, ATRIA, ORBIT). Methods: A nationwide cohort study of 40,450 AF patients who received warfarin for stroke prevention was performed. The clinical endpoints included ICH and major bleeding. Bleeding scores were compared using receiver operating characteristic (ROC) curves (areas under the ROC curves [AUCs], or c-index) and the net reclassification index (NRI). Results: During a follow up of 4.60 ± 3.62 years, 1581 (3.91%) patients sustained ICH and 6889 (17.03%) patients sustained major bleeding events. All tested bleeding risk scores at baseline were higher in those sustaining major bleeds. When compared to no ICH, patients sustaining ICH had higher baseline HEMORR2HAGES (p = 0.003), HAS-BLED (p < 0.001) and MBR factors score (p = 0.013) but not ATRIA and ORBIT scores. When HAS-BLED was compared to other bleeding scores, c-indexes were significantly higher compared to MBR factors (p < 0.001) and ORBIT (p = 0.05) scores for major bleeding. C-indexes for the MBR factors score was significantly lower compared to all other scores (De long test, all p < 0.001). When NRI was performed, HAS-BLED outperformed all other bleeding risk scores for major bleeding (all p < 0.001). C-indexes for ATRIA and ORBIT scores suggested no significant prediction for ICH. Conclusion: All contemporary bleeding risk scores had modest predictive value for predicting major bleeding but the best predictive value and NRI was found for the HAS-BLED score. Simply depending on modifiable bleeding risk factors had suboptimal predictive value for the prediction of major bleeding in AF patients, when compared to the HAS-BLED score.

Original languageEnglish
Pages (from-to)157-161
Number of pages5
JournalInternational Journal of Cardiology
Volume254
DOIs
StatePublished - 1 Mar 2018

Keywords

  • Atrial fibrillation
  • Bleeding
  • HAS-BLED
  • Risk factors

Fingerprint

Dive into the research topics of 'Major bleeding and intracranial hemorrhage risk prediction in patients with atrial fibrillation: Attention to modifiable bleeding risk factors or use of a bleeding risk stratification score? A nationwide cohort study'. Together they form a unique fingerprint.

Cite this