Patients who have a minor bleed during treatment with vitamin K antagonists (VKAs) have a 3-fold increased risk of subsequent major bleeding. The nature of the underlying risk factors is largely unknown.Objectives:
To indicate why patients with minor bleeds are at increased risk of subsequent major bleeds (e.g. are risk factors of a transient or a fixed nature).Methods:
Patients who started VKA treatment between 2003 and 2013 were included. Exposure was from the minor bleed until 3 months later. We used two analyses: a Cox model which we adjusted for several known risk factors, and a case-crossover (CCO) design, which corrects for all fixed risk factors (such as chronic diseases and genes) as patients are compared with themselves. The combination of both analyses gives insight into whether the association of minor with major bleeds is a result of fixed or transient risk factors.Results:
Out of 26 130 patients who were included and followed for ‘61 672 patient years’, 7194 experienced a minor bleed and 913 a major bleed. The Cox model indicated that patients with minor bleeds had a 2.5-fold increased risk of experiencing subsequent major bleeding after adjustment for known risk factors, whereas the CCO gave risk estimates around unity (odds ratio, 0.9; 95% confidence interval, 0.5–1.5).Conclusions:
The combination of both analyses indicates that minor bleeds are markers for fixed and currently unknown risk factors for major bleeding events.