Abstract
BACKGROUND: Oral anticoagulant drugs are proven to prevent thromboembolism in patients with atrial fibrillation (AF). To date, HAS-BLED score is used to assess bleeding risk. This study was conducted to compare simplified HAS-BLED (sHAS-BLED) with conventional HAS-BLED (cHAS-BLED) scores. METHODS: This retrospective study recruited patients with AF receiving warfarin among July 2013 to December 2018 in Central Chest Institute of Thailand. The cHAS-BLED score used the time in therapeutic range less than 70% as labile INR, whereas sHAS-BLED score used SAMe-TT(2)R(2) score of 3 or more as a substitute for labile INR. A paired Student's t test was used to compare sHAS-BLED and cHAS-BLED. The Pearson's correlation was used to assess the correlation of sHAS-BLED to cHAS-BLED scores. The Bland-Altman plot was used to confirm the agreement of individual sHAS-BLED to cHAS-BLED score. RESULTS: A total of 126 AF patients were enrolled. The average age, SAMe-TT(2)R(2) score, and cHAS-BLED score were 70.52 ± 10.37 years, 3.53 ± 1.03, and 2.03 ± 0.95, respectively. The sHAS-BLED score was not statistically significantly different compared with cHAS-BLED score (P = .08). The sHAS-BLED and cHAS-BLED scores had a very strong correlation with a correlation coefficient of .86 (P < .01). The Bland-Altman plot was performed to confirm the agreement of individual sHAS-BLED to cHAS-BLED scores. CONCLUSIONS: The sHAS-BLED was not statistically significantly different compared with cHAS-BLED and can be used in clinical practice. However, larger clinical trial will be needed to prove whether sHAS-BLED can predict bleeding risk in the future.