We were pleased to read the letter to the Editor written by Esteve-Pastor et al. and we thank them for their comments on our article.1 After a careful reading and interpretation of the letter, we would like to clarify some points.
As we know, bleeding risk assessment is more complex than thromboembolic risk assessment and every effort to improve bleeding risk assessment is welcomed. In our study, we found that ORBIT performed as well as HAS-BLED.1
It has been estimated that 90% of the deaths from vitamin K antagonist (VKA)-associated hemorrhage may occur within the first 30 days after the initiation of warfarin therapy (ie, in the period when we do not have enough data about international normalized ratio [INR] control).2 This in turn gives rise to continuous confusion about the significance of the labile INR element in the HAS-BLED score as this element is usually absent when bleeding risk is estimated in VKA-naïve patients (ie, the usual scenario in real world practice).
In the analysis of FANTASIIA,3 poor anticoagulation control (ie, labile INR) was defined as an estimated time in therapeutic range (TTR) < 65%, while in our study, we calculated labile INR as TTR < 60%. Therefore, any comparison between the 2 studies might be misleading.
HAS-BLED has several advantages (eg, it includes modifiable risk factors). However, HAS-BLED is composed of 9 elements, while ORBID is composed of just 5, which could explain why a higher percentage of patients are classified as having a high bleeding risk in HAS-BLED (20%-40%) than in ORBIT (5%-12%).3 This could result in an unnecessary delay in prescribing oral anticoagulants by junior physicians or inexperienced cardiologists who are not aware that high bleeding risk does not necessarily contraindicate anticoagulation. We believe that, as we will continue to use HAS-BLED, more efforts are needed to increase awareness among physicians of the proper use of this score, particularly regarding the high risk category.