Factors associated with abrupt discontinuation of dabigatran therapy in patients with atrial fibrillation in Malaysia.

Background Oral anticoagulant therapy is indicated for the prevention of stroke or other thromboembolic events. Premature discontinuation of oral anticoagulants may increase the risk of thromboembolism resulting in adverse sequelae. There are sparse data on the prevalence and the predictors of dabigatran discontinuation in Malaysian patients with atrial fibrillation. Objectives Determine the reasons and identify associated factors for abrupt discontinuation of dabigatran, assess the switching pattern and the occurrence of thromboembolic events after dabigatran discontinuation. Setting A university-affiliated tertiary hospital in Kuala Lumpur, Malaysia. Methods The clinical and demographic data of a cohort who were initiated with dabigatran between 2010 and 2012 at the University of Malaya Medical Centre were reviewed until the date of death or on 31st December 2013. Those patients who discontinued dabigatran were further followed up until 31st December 2015 to determine the occurrence of any thromboembolic event. Main outcome measure Permanent discontinuation of dabigatran for more than 8 weeks. Results 26 (14 %) of a cohort of 192 patients discontinued dabigatran therapy during a median follow-up period of 20 (range 3-45) months. About one-half of the discontinuation occurred within the first 6 months of dabigatran use. The three most cited reasons for discontinuation are bleeding events (19 %), high out-of-pocket drug payment (19 %) and cardioversion (19 %). Heart failure [adjusted odds ratio 3.699 (95 % confidence interval 1.393-9.574)] or chronic kidney disease [adjusted odds ratio 5.211 (95 % confidence interval 1.068-23.475)] were found to be independent risk factors for abrupt dabigatran discontinuation. Patients who discontinued dabigatran received warfarin (38 %), antiplatelet agents (16 %) or no alternative antithrombotic therapy (46 %). Five of the 26 patients who discontinued dabigatran developed an ischaemic stroke within 3-34 months after discontinuation. Conclusion Abrupt dabigatran discontinuation without an alternative oral anticoagulant increases the risk of thromboembolic events. As adverse drug events and renal impairment contribute substantially to the premature discontinuation of dabigatran, it is important to identify and monitor patients at risk to reduce dabigatran discontinuation rate especially during the first six months of dabigatran therapy.