Outcomes in VKA-treated patients with atrial fibrillation and chronic kidney disease: Clinical trials vs 'real-world'

Background Our objectives were to evaluate the risk of adverse events in a 'real-world' vs 'clinical trial' cohort of atrial fibrillation (AF) patients with chronic kidney disease (CKD). Methods We studied patient-level data for vitamin K antagonist-treated AF patients with a creatinine clearance <60 mL/min from the Murcia AF Project and AMADEUS trial. The study end-points were ischaemic stroke, major bleeding, all-cause mortality, myocardial infarction and intracranial haemorrhage. Results This study included 1,108 AF patients with CKD. The annual rate of the composite study outcome of ischaemic stroke, major bleeding and all-cause mortality was higher in the real-world (13.4%) vs AMADEUS (6.6%) cohort with an IRR of 2.04 (95% CI,1.34-3.09), P < .001. Individual annual rates of major bleeding, all-cause mortality and non-cardiovascular mortality were significantly greater in the real-world cohort. Similar findings were demonstrated even after multivariable adjustment, with the composite outcome HR of 2.85 (95% CI,1.74-4.66), P < .001. In a propensity score matched cohort, this risk remained significantly higher in the real-world cohort (IRR 2.95 [95% CI,1.03-10.28], P = .027), as did the risk of major bleeding and all-cause mortality. Conclusion Vitamin K antagonist-treated AF patients with CKD are exposed to significant annual rates of major adverse events including all-cause mortality. This risk may be under-appreciated in the idealised environment of randomised controlled trials.