Comparative assessment of the HAS-BLED score with other published bleeding risk scoring schemes, for intracranial haemorrhage risk in a non-atrial fibrillation population: the Chin-Shan Community Cohort Study.

Background: The HAS-BLED score is a validated bleeding risk model for predicting major bleeding events in anticoagulated individuals with atrial fibrillation (AF). It remains uncertain whether the HAS-BLED score could identify non-AF individuals at risk of developing intracranial haemorrhage (ICH), which is the most intractable and devastating major bleeding complication.

Methods: We assessed the predictive value of a modified HAS-BLED and other bleeding risk scoring models to predict the risk for ICH in the Chin-Shan Community Cohort, which followed 1899 women and 1703 men, aged >35 years, for a median of 15.9 years. ICH events (including haemorrhagic strokes) were ascertained according to questionnaires and the national register database.

Results: Of 3524 individuals without baseline AF, 54 ICH events occurred during follow-up. The risk for ICH was raised with increasing HAS-BLED scores, and was significantly associated with uncontrolled hypertension and older age (Odds Ratios [95% confidence interval (CI)], 4.2[2.3-7.6] and 1.9[1.1-3.4], respectively). Among the five bleeding risk scoring schemes tested, HAS-BLED had highest general discrimination performance (c-statistic [95% CI], 0.72 [0.67-0.78]), and better ability to discriminate between those who were at risk for ICH and who were not (NRI, net reclassification improvement, all p<0.05, compared to other four scoring schemes).

Conclusion: The HAS-BLED score had the highest general discrimination performance and best ability to discriminate risk for ICH. This score may be of clinical use in predicting the risk for occurrence of ICH among non-AF individuals.